00Harrison_FM.

qxd

2/25/05

12:38 PM

Page i

PRAISE

FOR

LINUX QUICK FIX NOTEBOOK

“Where has this book been all my professional life? It is about time someone wrote this book. The author covers many basic, practical issues that other authors omit in a format that makes for a great troubleshooting reference and for casual perusing. There are a lot of books available that give a broad overview of system and network administration, and there are a lot of detailed books on each specific topic covered in this book, but I don’t know of any other place where I can get so much useful, practical information on such a wide range of topics in such a concise and readable form. I will highly recommend it to my friends and colleagues who are trying to manage their own systems. I am anxious to show it to our instructors who teach data communications and networks as a possible text or lab manual.” —Ron McCarty, instructor and program chair, Computer Science, Penn State Erie “This book takes a very unique approach and is very flexible. It offers a great deal to the home user, students, small businesses—all the way to any Linux sys admin that needs a handy reference. I found it very refreshing in terms of style, stated goals, presentation, ease of use, direction, and the “present and future” value of the content. In other words the author is very current and knowledgeable. I would recommend this book to every level from programmer to manager.” —Joe Brazeal, Information Technician III, Southwest Power Pool “This book is an excellent resource for new Linux administrators. I was impressed with the comprehensive set of 35 topic chapters, making this book a very useful reference. I know of no other work like this.” —George Vish II, Linux curriculum program manager and senior education consultant, Hewlett-Packard Education “I would recommend this book to novice Linux users and administrators. It covers a lot of topics, and makes it easy for readers to follow along.” —Bret Strong, education consultant, Hewlett-Packard Company “Linux Quick Fix Notebook is a well-written guide to designing, installing, and maintaining Linux servers. Many Linux reference texts are just simple rewrites of Linux ¨man¨ and ¨info¨ pages, but this book addresses the real issues that systems administrators encounter when working with Linux. The author provides a pedagogical guide to learning Linux server administration. In addition, the book includes step-by-step information so that users can quickly take advantage of the best that open source software has to offer today. I would recommend this book to both beginning and intermediate users of Linux.” —Joseph Naberhaus, Professor of Networking, Universidad Del Mar

00Harrison_FM.qxd

2/25/05

12:38 PM

Page ii

00Harrison_FM.qxd

2/25/05

12:38 PM

Page iii

Linux Quick Fix Notebook
®

00Harrison_FM.qxd

2/25/05

12:38 PM

Page iv

BRUCE PERENS’ OPEN SOURCE SERIES
http://www.phptr.com/perens

◆ C++ GUI Programming with Qt 3

Jasmin Blanchette, Mark Summerfield

◆ Managing Linux Systems with Webmin: System Administration and

Module Development Jamie Cameron Mel Gorman

◆ Understanding the Linux Virtual Memory Manager ◆ Implementing CIFS: The Common Internet File System

Christopher Hertel

◆ Embedded Software Development with eCos

Anthony Massa

◆ Rapid Application Development with Mozilla

Nigel McFarlane

◆ The Linux Development Platform: Configuring, Using, and Maintaining a

Complete Programming Environment Rafeeq Ur Rehman, Christopher Paul

◆ Intrusion Detection with SNORT: Advanced IDS Techniques Using SNORT,

Apache, MySQL, PHP, and ACID Rafeeq Ur Rehman

◆ The Official Samba-3 HOWTO and Reference Guide

John H. Terpstra, Jelmer R. Vernooij, Editors

◆ Samba-3 by Example: Practical Exercises to Successful Deployment

John H. Terpstra

00Harrison_FM.qxd

2/25/05

12:39 PM

Page v

Linux Quick Fix Notebook
®

Peter Harrison

Upper Saddle River, NJ • Boston • Indianapolis • San Francisco New York • Toronto • Montreal • London • Munich • Paris Madrid • Capetown • Sydney • Tokyo • Singapore • Mexico City

00Harrison_FM.qxd

3/1/05

4:08 PM

Page vi

Many of the designations used by manufacturers and sellers to distinguish their products are claimed as trademarks. Where those designations appear in this book, and the publisher was aware of a trademark claim, the designations have been printed with initial capital letters or in all capitals. The author and publisher have taken care in the preparation of this book, but make no expressed or implied warranty of any kind and assume no responsibility for errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of the use of the information or programs contained herein. The publisher offers excellent discounts on this book when ordered in quantity for bulk purchases or special sales, which may include electronic versions and/or custom covers and content particular to your business, training goals, marketing focus, and branding interests. For more information, please contact: U. S. Corporate and Government Sales (800) 382-3419 corpsales@pearsontechgroup.com For sales outside the U. S., please contact: International Sales international@pearsoned.com Visit us on the Web: www.phptr.com Library of Congress Number: 2004117224 Copyright © 2005 Pearson Education, Inc. Printed in the United States of America. This material may be distributed only subject to the terms and conditions set forth in the Open Publication License, v1.0 or later (the latest version is presently available at http://www.opencontent.org/openpub/). ISBN 0-13-186150-6 Text printed in the United States on recycled paper at R.R. Donnelley and Sons Company in Crawfordsville, Indiana. First printing, March 2005

00Harrison_FM.qxd

2/25/05

12:39 PM

Page vii

Dedicated to Diana “Turn off the light and come to bed.” Blaze and Jade “Daddy, can we stop being quiet now?” Jack and Barbara By my side every day.

00Harrison_FM.qxd

2/25/05

12:39 PM

Page viii

00Harrison_FM.qxd

2/25/05

12:39 PM

Page ix

Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xxxi Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xl About the Author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xli

Part 1

The Linux File Server Project

1 Why Host Your Own Site? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 Our Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 Alternatives to In-House Web Hosting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 Virtual Hosting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 Dedicated Hosting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 Factors to Consider Before Hosting Yourself . . . . . . . . . . . . . . . . . . . . . . . . . .6 Is In-House Hosting Preferred? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 How to Migrate From an External Provider . . . . . . . . . . . . . . . . . . . . . . . . . . .8 In-House Server Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10 2 Introduction to Networking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 The OSI Networking Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12 An Introduction to TCP/IP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13 TCP Is a Connection-Oriented Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . .14 How TCP Establishes a Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14 UDP, TCP’s “Connectionless” Cousin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 TCP and UDP Ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 The TCP/IP Time to Live Feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16 The ICMP Protocol and Its Relationship to TCP/IP . . . . . . . . . . . . . . . . . .16

ix

00Harrison_FM.qxd

2/25/05

12:39 PM

Page x

x

Contents

How IP Addresses Are Used to Access Network Devices . . . . . . . . . . . . . . .17 Private IP Addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17 The localhost IP Address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18 Network Address Translation (NAT) Makes Private IPs Public . . . . . . . .18 Port Forwarding with NAT Facilitates Home-Based Web Sites . . . . . . . .19 DHCP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19 How DNS Links Your IP Address to Your Web Domain . . . . . . . . . . . . . . .20 IP Version 6 (IPv6) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21 How Subnet Masks Group IP Addresses into Networks . . . . . . . . . . . . . . . .21 Calculating the Number of Addresses Assigned to a Subnet . . . . . . . . . . .22 Calculating the Range of Addresses on Your Network . . . . . . . . . . . . . . . .22 Subnet Masks for the Typical Business DSL Line . . . . . . . . . . . . . . . . . . .24 The Physical and Link Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24 Networking Equipment Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25 Network Interface Cards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25 The Two Broad Types of Networking Equipment . . . . . . . . . . . . . . . . . . .27 Connectivity Using Hubs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28 Using Switches as a Faster Alternative to Hubs . . . . . . . . . . . . . . . . . . . .29 Local Area Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29 How Routers Interconnect LANs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29 How Simple Routing Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30 Default Gateways, the Routers of Last Resort . . . . . . . . . . . . . . . . . . . . . .31 Firewalls Help Provide a Secure Routing Environment . . . . . . . . . . . . . .31 Additional Introductory Topics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32 The File Transfer Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32 Linux Help . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33 3 Linux Networking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35 How to Configure Your NIC’s IP Address . . . . . . . . . . . . . . . . . . . . . . . . . . . .36 Determining Your IP Address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36 Changing Your IP Address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37 Multiple IP Addresses on a Single NIC . . . . . . . . . . . . . . . . . . . . . . . . . . .39 IP Address Assignment for a Direct DSL Connection . . . . . . . . . . . . . . . .40 IP Address Assignment for a Cable Modem Connection . . . . . . . . . . . . . .45 How to Activate/Shut Down Your NIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45 How to View Your Current Routing Table . . . . . . . . . . . . . . . . . . . . . . . . . . . .45 How to Change Your Default Gateway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .46 How to Configure Two Gateways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47 Adding Routes from the Command Line . . . . . . . . . . . . . . . . . . . . . . . . . . .47 Adding Routes with /etc/sysconfig/network-scripts/ Files . . . . . . . . . . .48 How to Delete a Route . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49 Changing NIC Speed and Duplex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49 Using mii-tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49 Using ethtool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50

00Harrison_FM.qxd

2/25/05

12:39 PM

Page xi

Contents

xi

How to Convert Your Linux Server into a Router . . . . . . . . . . . . . . . . . . . . . .52 Configuring IP Forwarding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .52 Configuring Proxy ARP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53 Configuring Your /etc/hosts File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .54 The loopback Interface’s localhost Entry . . . . . . . . . . . . . . . . . . . . . . . . . .55 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56 4 Simple Network Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57 Sources of Network Slowness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58 Sources of a Lack of Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58 Doing Basic Cable and Link Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58 Testing Your NIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59 Viewing Your Activated Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59 Viewing All Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59 Testing Link Status from the Command Line . . . . . . . . . . . . . . . . . . . . . .60 Viewing NIC Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .61 How to See MAC Addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62 Using ping to Test Network Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . .64 Using telnet to Test Network Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . .65 Linux telnet Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66 Successful Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66 Connection Refused Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .67 telnet Timeout or Hanging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .67 telnet Troubleshooting Using Windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . .67 Screen Goes Blank—Successful Connection . . . . . . . . . . . . . . . . . . . . . . . .68 Connect Failed Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .68 telnet Timeout or Hanging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .68 Testing Web Sites with the curl and wget Utilities . . . . . . . . . . . . . . . . . . . . .68 Using curl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .68 Using wget . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69 The netstat Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70 The Linux iptables Firewall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 Determining Whether iptables Is Running . . . . . . . . . . . . . . . . . . . . . . . .71 How to Stop iptables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 How to Configure iptables Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 Using traceroute to Test Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 Sample traceroute Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72 Possible traceroute Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72 traceroute Time Exceeded False Alarms . . . . . . . . . . . . . . . . . . . . . . . . . . .73 traceroute Internet Slowness False Alarm . . . . . . . . . . . . . . . . . . . . . . . . .74 traceroute Dies at the Router Just Before the Server . . . . . . . . . . . . . . . .75 Always Get a Bidirectional traceroute . . . . . . . . . . . . . . . . . . . . . . . . . . . .75 ping and traceroute Troubleshooting Example . . . . . . . . . . . . . . . . . . . . . .77 traceroute Web Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77 Possible Reasons for a Failed traceroute . . . . . . . . . . . . . . . . . . . . . . . . . .78

00Harrison_FM.qxd

2/25/05

12:39 PM

Page xii

xii

Contents

Using MTR to Detect Network Congestion . . . . . . . . . . . . . . . . . . . . . . . . . . .79 Viewing Packet Flows with tcpdump . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79 Analyzing tcpdump files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82 Common Problems with tcpdump . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82 Viewing Packet Flows with tethereal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 Basic DNS Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85 Using nslookup to Test DNS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85 Using the host Command to Test DNS . . . . . . . . . . . . . . . . . . . . . . . . . . . .86 Using nmap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87 Determining the Source of an Attack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88 Who Has Used My System? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 The last Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 The who Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90 5 Troubleshooting Linux with syslog . . . . . . . . . . . . . . . . . . . . . . . . . . . .91 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .91 syslog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 The /etc/syslog.conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 Activating Changes to the syslog Configuration File . . . . . . . . . . . . . . . .93 How to View New Log Entries as They Happen . . . . . . . . . . . . . . . . . . . . .93 Logging syslog Messages to a Remote Linux Server . . . . . . . . . . . . . . . . .94 syslog Configuration and Cisco Network Devices . . . . . . . . . . . . . . . . . . .96 syslog and Firewalls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96 logrotate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96 The /etc/logrotate.conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96 Sample Contents of /etc/logrotate.conf . . . . . . . . . . . . . . . . . . . . . . . . . .97 The /etc/logrotate.d Directory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 Activating logrotate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98 Compressing Your Log Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .99 6 Installing RPM Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101 Where to Get Commonly Used RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102 RPMs on Your Installation CDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102 RPMs Downloaded from Fedora . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102 RPMs Downloaded from rpmfind.net . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 Getting RPMs Using Web-Based FTP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 Using the Fedora Web site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 Using the rpmfind Web site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 Getting RPMs Using Command-Line Anonymous FTP . . . . . . . . . . . . . . . .104 Getting RPMs Using wget . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .107 Automatic Updates with yum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .107 Configuring yum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .108 Creating Your Own yum Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .109

00Harrison_FM.qxd

2/25/05

12:39 PM

Page xiii

Contents

xiii

Before You Start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .110 Keeping Your System Current with yum . . . . . . . . . . . . . . . . . . . . . . . . . .110 Example of a yum Package Installation . . . . . . . . . . . . . . . . . . . . . . . . . . .112 How to Install RPMs Manually . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Using Downloaded Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Using CD-ROMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 How to Install Source RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .114 RPM Installation Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115 Failed Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115 Signature Keys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .116 How to List Installed RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117 Listing Files Associated with RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117 Listing Files for Already Installed RPMs . . . . . . . . . . . . . . . . . . . . . . . . .118 Listing Files in RPM Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118 Listing the RPM to Which a File Belongs . . . . . . . . . . . . . . . . . . . . . . . . .119 Uninstalling RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Which RPMs Start at Boot Time? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Installing Software Using tar Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .120 7 The Linux Boot Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .121 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .121 The Fedora Boot Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .121 Determining the Default Boot runlevel . . . . . . . . . . . . . . . . . . . . . . . . . . . . .123 Getting a GUI Console . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124 Get a Basic Text Terminal Without Exiting the GUI . . . . . . . . . . . . . . . . . .125 Using a GUI Terminal Window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125 Using Virtual Consoles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125 System Shutdown and Rebooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125 Halt/Shut Down the System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125 Reboot the System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .126 Entering Single-user Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .127 How to Set Programs to Run at Each runlevel . . . . . . . . . . . . . . . . . . . . . . .129 chkconfig Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .129 Using chkconfig to Improve Security . . . . . . . . . . . . . . . . . . . . . . . . . . . .130 Final Tips on chkconfig . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .132 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .133 8 Configuring the DHCP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .135 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .135 Download and Install the DHCP Package . . . . . . . . . . . . . . . . . . . . . . . . . .135 The /etc/dhcpd.conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .136 How to Get DHCP Started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .138 DHCP Servers with Multiple NICs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .138 Temporary Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139 Permanent Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139

. . . .163 Create Samba Trust Accounts for Each Workstation . . . . . . . . . . . . .145 Simple /etc/sudoers Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .142 How to Delete Users . . . . . . . .149 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . .162 Home Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141 Adding Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . .142 How to Change Passwords . . . . . . . . . . . . . . . .158 The [global] Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .150 How to Get Samba Started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Linux. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141 In This Chapter . . . . . .151 How SWAT Makes Samba Simpler . . . . .158 Creating a Starter Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .140 Conclusion . . . . . . . . . . . . . . . . . . .154 How to Make SWAT Changes Immediate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139 DHCP Clients Obtaining 169. . . . .144 Example of a User Using sudo . . . . . . . . . . . . . . . .146 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .158 The [homes] Section . . . . . . . . . . . . . . . . . . .154 Encrypting SWAT . . . .qxd 2/25/05 12:39 PM Page xiv xiv Contents Configuring Linux Clients to Use DHCP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139 Configuring Windows Clients to Use DHCP . . . . . . . . . . . . . .162 How to Create a Samba PDC Administrator User . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .0 Addresses . . . . . . . . . . . . . . . . . . .161 Shares for Specific Groups of Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163 How to Add Workstations to Your Samba Domain . . . . . . .141 How to Add Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141 Who Is the Super User? . . . . .139 Simple DHCP Troubleshooting . . . . . . . . . . . . . . . . . . . . . . and Samba . . . . . . . . . . . . . . . . . . . . . . . . . .163 Make Your PC Clients Aware of Your Samba PDC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .152 Controlling SWAT . . . . . . .144 Downloading and Installing the sudo Package .254.144 Using sudo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162 Corporate Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162 Samba Passwords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .152 Basic SWAT Setup . . .140 9 Linux Users and sudo . . . . . . . . . . . . . . . .160 The [netlogon] and [profiles] Share Sections . . . . . . . . . . . . . . . . . .149 Download and Install Packages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .143 How to Change the Ownership of a File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .148 10 Windows. . .151 The Samba Configuration File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161 The [printers] Share Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .143 How to Tell the Groups to Which a User Belongs . . .0. .164 . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189 Encryption Keys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .166 Mapping a Private Windows Drive Share . . . . . . .168 How to Delete Users from Your Samba Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . .171 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171 Make Samba Aware of the Printer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .182 Checking the Samba Logs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186 Samba Network Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xv Contents xv How to Add Users to Your Samba Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .166 Mapping the Linux Users to an smbpassword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186 Basic Samba Security . . . .179 In This Chapter . . . . . . . . . . . . . . . . . .188 Common Wireless Networking Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . .181 Testing Basic Client/Server Network Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .173 Creating Group Shares in Samba . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . .189 . . .179 Samba and Firewall Software . . . . . . . . . . . . . . . . . . . . . . . . .166 Adding the Users in Linux . . . . . . . .175 Windows Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186 Conclusion .168 How to Modify Samba Passwords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .181 Testing Samba Client/Server Network Connectivity . . . . .177 Automating Mounting with Linux Samba Clients . . . . . . . . . . . .188 Wireless Access Points . . . . . . . . .174 Configure the Share in SWAT . . . . . . . . . . . . . .173 Configure the Printer Driver on the Workstations . . . . . . . . . . . . . . . . . . . . . . . . . . . .187 Wireless Linux-Compatible NICs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167 Domain Groups and Samba . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .179 Testing the smb. . . . . . . . . . . . . . . . . .180 Windows-based Zone Alarm .187 In This Chapter . . . . . . . . . . . . . . .174 Create the Directory and User Group . . . . .178 12 Samba Security and Troubleshooting . . . . . . . . . . . . . . . . .176 Create a CD-ROM Drive Mount Point on Your Samba Server . . . . . . . . . . . . . . . .188 Extended Service Set ID . . . . . . . . . . . . . . . . . . . . . . . . . . .171 Adding a Printer to a Samba PDC . . . .181 The Windows XP Built-In Firewall . . . .186 13 Linux Wireless Networking . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171 Adding the Printer to Linux . . . . . . . . . . . . . . . . . . . . . . . . . . . . .180 Linux iptables . . . . . . .175 Test Your Windows Client Configuration .169 11 Sharing Resources Using Samba . . . . . .178 Conclusion . . . . . . . .174 Map the Directory Using My Computer . .conf File . . .175 Sharing Windows Drives Using a Linux Samba Client . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .223 Basic iptables Initialization . . . . . . . . . . . . . . . . . . . . .199 Check the NIC Status . . . . . . . . . . . . .203 Can’t Ping Default Gateway . . . . . . . . . . . . . . . . . . . .210 Download and Install the iptables Package . . . . . . . .189 Permanent Wireless Tools Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .196 Linux-WLAN Encryption for Security . . . . . . . . . .227 Allowing WWW and SSH Access to Your Firewall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .206 Part 2 The Linux Web Site Project 14 Linux Firewalls Using iptables . . . . . . . . . . . . .222 Sample iptables Scripts . . . . . . . . .209 What Is iptables? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .200 Check for Interrupt Conflicts . . . . . . . . . . . . . . . .191 Linux-WLAN Preparation . . . . . . . . . . . . . . . . . . . . . . . . . .220 Fedora’s iptables Rule Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .216 Using User-Defined Chains . . . . . . .202 Kernel Log Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .226 Allowing DNS Access to Your Firewall . . . . . . . .209 In This Chapter .193 Linux-WLAN Post Installation Steps . . . . . . . . . . . . . . . . . . . . . . . . .205 Conclusion .197 Troubleshooting Your Wireless LAN . . . . . . .225 Advanced iptables Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .211 How to Start iptables . . . .215 Important iptables Command Switch Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191 Networking with Linux-WLAN . . . . . . . . . . . . . . . . . . .219 Saving Your iptables Scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .196 Testing Linux-WLAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .211 Determining the Status of iptables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .220 Recovering from a Lost Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .227 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xvi xvi Contents Networking with Linux Wireless Tools . .192 Installing the RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .190 Wireless Tools Encryption . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .221 Loading Kernel Modules Needed by iptables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .192 Identifying the Correct RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .203 Unknown Device Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .204 Wireless Networks In Businesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .211 Packet Processing in iptables . . . . . . . . . . . . . . . . . . . . . . .194 Start Linux-WLAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .203 A Common Problem with Linux-WLAN and Fedora Core 1 . . . . . . .211 Targets and Jumps . . . . . . . . . . . . . . . . . . . . . . . . .189 Using iwconfig for Wireless Tools Configuration . . . . . . . . . . . . . . . .222 Basic Operating System Defense . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . TFTP. . . . . . . . . . . . . . . . . . . . . . . . . . . . .252 Using The TELNET Client . . . . . . . . . .252 Installing the TELNET Server Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .251 Controlling xinetd-Managed Applications . . . . . . . .230 Static NAT . . . . . . . . . .conf Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .248 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .246 FTP Users with Read-Only Access to a Shared Directory . . . . . . . . . . . . . . . .232 Troubleshooting iptables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .242 Testing the Status of VSFTPD . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . .252 TELNET . .246 Troubleshooting FTP . . . . . . . . . . . . . . . . . . . . . . . . . . . .246 Tutorial . . . . . . . . . . . . .240 Server Protected by a Firewall Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .242 How to Get VSFTPD Started . . . . . . . . . . . . . . . . . .234 iptables Won’t Start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .236 15 Linux FTP Server Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .228 Allow Your Home Network to Access the Firewall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .234 Checking the Firewall Logs . . . . . . . . . . . . . . . . .245 FTP Greeting Banner . . . . . . . . . .243 The vsftpd. . . . . . . . . . . . . . . . . .conf File . . . . . . . .251 Managing xinetd Programs . . . . . . . . . . . . . . . . .254 Stopping a TELNET Server . . . . . . . . . . . . . . . . . . . . . . . . .240 Client Protected by a Firewall Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .238 Problems with FTP and Firewalls . . . .qxd 2/25/05 12:39 PM Page xvii Contents xvii Allowing Your Firewall to Access the Internet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .253 Setting Up a TELNET Server . . . . . . . . . . . . . . . . .236 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .237 In This Chapter . . . . . . . . . . . . . . . . .245 FTP Security Issues . . . . . . . . . . . . . . . . . . . . .251 In This Chapter . . . . . . . . . .237 FTP Overview . . . . . . . . . . . . . . . . . . . . . . . . . . .246 Using SCP as Secure Alternative to FTP . . . . . . . . . . . . . . . . . . . . . . . . . .237 Types of FTP . . . .229 Masquerading (Many to One NAT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .229 Port Forwarding Type NAT (DHCP DSL) . . . . . . .254 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .249 16 Telnet. . . . . . . . . . . . . . . . . . . . .254 Basic TELNET Security . . . . . . . . . .245 The /etc/vsftpd. . . . .247 Sample Login Session to Test Functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . and xinetd . . . . . . . . .241 How to Download and Install VSFTPD . . . . . . . . .243 Other vsftpd. . . . . . .245 Anonymous Upload . . .ftpusers File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .251 Controlling xinetd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.277 When To Use A Dynamic DNS Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .271 Using SSH and SCP Without a Password . . . . . . . . . . . . . . . . . . . . . . . . . . .279 . . . . . . . . . . . . . .273 18 Configuring DNS . . . . . . . . . . . . . . . .266 Using SSH to Log Into a Remote Machine . . . . . . . . . . . . . . . . . . . . . . . .262 17 Secure Remote Logins and File Copying . . . . . . . . . . . . . . . . . . . . . . . . .265 Testing the Status of SSH . . . . . . . . .270 SFTP: Another Secure Alternative to FTP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .275 DNS Domains . . . . . . . .265 SSH Versions 1 and 2 . . . . . .276 When to Use a Static DNS Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .256 Installing the TFTP Server Software . . . . . . . . . . . . . . . . .265 The /etc/ssh/sshd_config File . . . . . . . .267 What to Expect with Your First Login . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .257 Configuring the TFTP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . .276 BIND . .263 In This Chapter . . . . .261 Conclusion . . . . . . . . . . . . . .269 Copying Files to the Local Linux Box . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .263 A Quick Introduction to SSH Encryption . . . . . . . . . . . . . . . . . . . . . . . .275 Introduction to DNS . . . . . . . . .269 SCP: A Secure Alternative to FTP . . . . . . . . . . . . . . . . . . .277 Basic DNS Testing of DNS Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .277 How to Get Your Own Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .271 Configuration: Client Side . . . . . . . . . . . . . .276 When to Use a DNS Caching Name Server . . .276 How DNS Servers Find Your Site Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . .268 Executing Remote Commands on Demand with SSH . . . . . . . . . . . . .257 Saving Cisco Configurations to the TFTP Server . . . . . . . . . . . . . .258 Uploading Cisco Configurations from the TFTP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xviii xviii Contents TFTP . . . . . . . . . . . .279 Starting BIND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .278 Downloading and Installing the BIND Packages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .268 Deactivating TELNET After Installing SSH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .270 Copying Files to the Remote Linux Box . . . . . . . . . . . . . . . . . . . . . .272 Configuration: Server Side . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .263 Starting OpenSSH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .266 Change the TCP Port on Which SSH Listens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .273 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .276 DNS Clients . . . . . . . . . . .276 Authoritative DNS Servers . . . .275 In This Chapter . . . . . . . . . . . . . . . . . .260 Using TFTP to Restore Your Router Configuration . .267 SSH Failures Due to Linux Reinstallations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .307 The /etc/ddclient. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .282 Configuring the Zone Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .292 Make Sure Your /etc/hosts File is Correctly Updated . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .305 Install a DDNS Client on Your Server .300 Zone Transfer Protection .282 Configuring named. . . . . . . . . . . . . . . . . . . . . . . . .310 Conclusion . . . . . . . . . . . .288 What You Need to Know About NAT and DNS . . . . . . . . . . . . . . . . . . . . . . .295 Migrating Your Web Site In-House . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .301 19 Dynamic DNS . .298 DHCP Considerations for DNS . . . . . . . . . . . . . . . . . . .307 DDNS Client Software: Linux DDclient . . . . . .299 Simple DNS Security . . . . . . . . . . .313 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .280 Important File Locations . . .307 DDNS Client Software: SOHO Router/Firewalls . . . . . .279 Configuring a Caching Name Server .294 General Troubleshooting Steps . . . . .313 Starting Apache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .313 Downloading and Installing the Apache Package . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .300 Naming Convention Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf File . . . . . . . . . . . . . . . . . . . .309 Finding DDclient Help . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .289 Loading Your New Configuration Files . . . . . . . . . . . . . . . . . . . . . .293 Fix Your Domain Registration . . . . . . . . . . . . . . . . . . . .310 Testing Your Dynamic DNS . . . . . . . . . . . . .292 Configure Your Firewall . . . . . .311 20 The Apache Web Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .284 Sample Forward Zone File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .281 Configuring a Regular Name Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .301 Conclusion . . . . . . . . . . . . . . .304 Testing ISP Connectivity for Your Web Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .287 Sample Reverse Zone File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .303 In This Chapter . . . . . .293 Troubleshooting BIND . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xix Contents xix The /etc/resolv. .282 Configuring resolv. . . . . . . .306 Dynamic DNS and NAT Router/Firewalls . . . . . . . . . . . . . . . . . . . . . . . .314 . . . . . . . . . . . . . .303 Dynamic DNS Preparation . . . . . . . . . . . . .314 DHCP and Apache . . .314 Configuring DNS for Apache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. .308 How to Get DDclient Started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .304 Registering DDNS . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . .328 Apache Running on a Server Behind a NAT Firewall . . . . . . . . . . . . . . . . . . . . . . . . . . .329 Step 1: Configure Virtual Hosting on Multiple IPs . . . .330 The /etc/httpd/conf. . . . . . . . . . . .331 Troubleshooting Apache . . . . . . . . . .335 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .329 Step 2: Configure DNS Views . .323 Testing Your Web Site Before DNS Is Fixed . . . . . . . . . . . . . . . . . . .332 Browser 403 Forbidden Messages . . . . . . . . . . .347 sendmail Masquerading Explained .333 The Apache Error Log Files . .347 Which User Should Really Receive the Mail? . . . . . . . . . . . . . . . . .conf Files When Upgrading . . . . . . .327 Using Data Compression on Web Pages . . . . . . . . .322 Configuration: Multiple Sites and IP Addresses . .337 Configuring sendmail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .329 Protecting Web Page Directories with Passwords . . . . . . . . . . . . . . . . . . . . .332 Testing Basic HTTP Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . .342 How to Configure Linux sendmail Clients . . . . . . . . . . .352 Troubleshooting sendmail . . . . .mc File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .345 The /etc/mail/access File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .346 The /etc/mail/local-host-names File . . . . . . . .332 Incompatible /etc/httpd/conf/http. . . . . . . . . . . . . .321 Virtual Hosting and SSL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .332 Only the Default Apache Page Appears . . . . . . . . . . . . .353 . . . .350 Using sendmail to Change the Sender’s E-mail Address . . . . . . . . . .327 Compression Configuration Example . . .333 The Apache Status Log Files . . . . . . .d Directory .320 IP-Based Virtual Hosting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .333 Server Name Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .340 Starting sendmail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .315 Where to Put Your Web Pages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .325 Disabling Directory Listings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .343 The /etc/mail/relay-domains File . . . . . .315 Named Virtual Hosting . . . . . . . . .340 The /etc/mail/sendmail. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .326 Handling Missing Pages .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .343 Converting From a Mail Client to a Mail Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . .341 Configuring DNS for sendmail . . . . . . .qxd 2/25/05 12:39 PM Page xx xx Contents General Configuration Steps . . . . . . . .340 How to Restart sendmail After Editing Your Configuration Files . . . . . . .335 21 Configuring Linux Mail Servers . . . . . . . . . . . . . . .337 How sendmail Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .337 Installing sendmail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .337 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .373 MRTG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .377 Configuring Apache to Work with MRTG . . . . . . . . . . . . . . . . . . . . . . . . . .365 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .383 23 Advanced MRTG for Linux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .382 Conclusion . . . . . . . . . . . . . . . . . . .389 . . . . .375 Getting MRTG to Poll Multiple Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .381 The vmstat Command . . . . . . . . . . .378 Troubleshooting MRTG . . . . . . . . . . . .362 How to Configure Your Windows Mail Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .358 spamassassin . . . . . . . . . . .365 OIDs and MIBs . . . . . . . . . . . . . . . . . . . . . . . . . . .363 Configuring Secure POP Mail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .365 SNMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .371 Simple SNMP Troubleshooting . . . . . . . . . . . . . .387 Differences in MIB and MRTG Terminology . . . . . . . . . . . . . . . .363 How to Handle Overlapping E-mail Addresses . . . . . . .380 How to View Your Webalizer Statistics .388 Manually Configuring Your MRTG File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .374 Configuring MRTG . .373 MRTG Download and Installation . . . . . . . . . . . . . . . . . . . . . . . . . .385 Testing Your MIB Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . .387 The CPU and the Memory Monitoring MIB . . . . . . . . . . . . . .382 The free Utility . . . . . . . . . . . . . . . . . . . . . . . . . .388 The TCP/IP Monitoring MIB . . . . . . . . . .365 SNMP Community Strings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .363 Troubleshooting POP Mail . . . . . . . . . . . . . . . . . . .358 Using Public Spam Blacklists with sendmail .369 Basic SNMP Security . . . . . . .359 A Simple PERL Script to Help Stop Spam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .364 22 Monitoring Server Performance . . . . . . . . . . . .385 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .368 SNMP Versions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .381 The top Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .377 How to View the MRTG Graphs in Your Web Browser . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .385 Locating and Viewing the Contents of Linux MIBs . . . . . . . . . . . . . . . . . . . . . . . . . .362 Starting Your POP Mail Server . . . . . .361 Configuring Your POP Mail Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .378 Using MRTG to Monitor Other Subsystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .379 Webalizer . . . . . . . . . . . . . . . . . .368 Doing SNMP Queries . . . . . . . . .381 The Webalizer Configuration File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .362 Installing Your POP Mail Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xxi Contents xxi Fighting Spam . . . .363 Conclusion . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .399 Testing The OIDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf File . . . . . . . . . . . . . . . . .399 Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .398 Configuring CRON to Use the New MRTG File . . . .407 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .395 A Complete Sample Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .390 Title Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .402 The /etc/ntp. . . . . . . . . . .412 Set Up Your Web Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .391 Scaling Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .395 Testing the Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . .411 In This Chapter . . . . .406 Cisco IOS . . . . . . . . . . . . . . . . . . . . . . . .395 Implementing Advanced Server Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .403 Doing an Initial Synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . .390 Options Parameters . . . .403 Determining If NTP Is Synchronized Properly . . . . . . . . . . . . . . . . . . . . . .404 Your Linux NTP Clients Cannot Synchronize Properly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .408 Part 3 Advanced Topics 25 Network-Based Linux Installation . . . . . . . . . . . . . . . . . .400 24 The NTP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .414 Create a Special FTP User .403 Verifying NTP is Running . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xxii xxii Contents Parameter Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .397 Creating a New MRTG Index Page to Include this File . . . . . . . . . . . . . . . . . .389 Legend Parameters . . . . . . . . . . . . . . . . . . . . . . . .400 Conclusion . . . . . . . . . . . . . . . . . . . . . . . .402 How to Get NTP Started . . . . .401 Download and Install the NTP Package . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .398 Monitoring Non-Linux MIB Values . . . . . . . . . .412 Basic Preparation . . . . . . . . . . . . . . . . . . . . . . .405 Fedora Core 2 File Permissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .406 CAT OS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .407 Configuring a Windows NTP Client . . . . . . . . . . . . . . . . . . . . .391 Defining The MIB Target Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .391 Defining Global Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .411 Setting Up the Installation Server . . . . . . . . . . .398 Scenario .407 Firewalls and NTP . . . . . . . . . . . . . . . . . . .403 Testing and Troubleshooting NTP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .415 . . . . . . .401 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . .406 Configuring Cisco Devices to Use an NTP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .398 Find the OIDs . . .414 Set Up Your FTP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .418 How to Run a Kickstart Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .435 Scenario . . .422 RAID 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .435 Adding Disks to Linux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .426 Configuring Software RAID . . . . . . . . . . . . . . .430 Create the RAID Set . . . .qxd 2/25/05 12:39 PM Page xxiii Contents xxiii Set Up Your NFS Server . . . . . . . . .416 Troubleshooting the Network Installation . . . . . . . . . .417 Automating Installation with Kickstart . . . . . . . . . . . . . . . . . . . . . . . . . . .416 Creating a Boot CD . . . . . . . . . . . . . . . . . . . . . . . .431 Load the RAID Driver for the New RAID Set . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . .435 Determining the Disk Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .431 Confirm RAID Is Correctly Initialized . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .432 Create a Mount Point for the RAID Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .424 Serial ATA Drives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .416 The Network Installation . . . . . . . . . . . . . . . . .432 Edit the /etc/fstab File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .423 RAID 5 . . .431 Format the New RAID Set .425 SCSI Drives . . . . . . . . . . . . . . . . . . . . . . . . . . .423 Before You Start . . . . . .425 Should I Use Software RAID Partitions or Entire Disks? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .421 RAID 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .436 . . . . . . . .415 Configure Your DHCP Server . . . . . .435 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .434 27 Expanding Disk Capacity . . . . . . . . . . . . . .421 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . .417 How to Create New Kickstart Configuration Files . . . .434 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . .419 Booting With Your Kickstart Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .427 RAID Partitioning . . . . . . . . . . . . . .433 Mount the New RAID Set . .420 26 Linux Software RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .423 RAID 4 . . . . . . . .417 Differences Between Fedora and Red Hat Installation . . . . . . . . . . . . . . . . . . .421 RAID Types . . . . . . . . . . . . . . . . . . .432 Start the New RAID Set’s Driver . . . . . . . . . . . . . . . . . . . .426 Back up Your System First . . . . . . . . . . . . . . .427 Edit the RAID Configuration File . . . . . . . . . .424 IDE Drives . . . . . . . . . . . . . . . . . . . . . . .426 Configure RAID in Single-User Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .419 Conclusion . . . .421 Linear Mode RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .433 Check the Status of the New RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

452 Remount the Filesystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . .462 Configuring NFS on the Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .451 Setting Up Quotas . . . . . . . . . . .463 Configuring NFS on the Client . . . . . .452 Get Out of Single-User Mode . . . . . . . . . . . .453 Initialize the Quota Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .455 29 Remote Disk Access With NFS . . . . . . . . . .465 Activating Modifications to the /etc/exports File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .442 LVM Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .457 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . or Modifying a Share . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .464 NFS and DNS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .461 Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xxiv xxiv Contents Preparing Partitions on New Disks . . . . . . . . . . . . . . . . . . . . . .462 Starting NFS on the Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .458 Key NFS Concepts . .454 Other Quota Topics . . . . . . . . . . .468 . . .463 Starting NFS on the Client . . . . .464 Making NFS Mounting Permanent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .467 Deleting. . . . . . . . . . . . . . . . . . . . . . . . . . . .440 Expanding Partitions with LVM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .443 Conclusion . . . . .451 Enter Single-User Mode . . Moving. . . . . . . . . . . . . . .451 Edit Your /etc/fstab File . . . . . . . . . . . . . . . . . . . . . . . . . .458 Important NFS Daemons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .442 Configuring LVM Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .462 The /etc/exports File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .457 General NFS Rules . . . . . . . . . . . . . . . . . . . . . . . . .455 Getting Quota Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .457 NFS Operation Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .437 Verifying Your New Partition . . . . . . . . . . . . . . . . . . .467 New Exports File . .453 Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .453 Edit the User’s Quota Information . . . . . . . . . . . . . . .455 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .451 In This Chapter .439 Putting a Directory Structure on Your New Partition . . .454 Editing Group Quotas . . . . . . . . . . . . . . .449 28 Managing Disk Usage with Quotas . . . . . . . . . . . . . . . . . . . .439 Migrating Data to Your New Partition . . . . . . . . . .467 Adding a Shared Directory To An Existing Exports File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .454 Editing Grace Periods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .460 Installing nfs . . . . . . . . . . . . . . . . . .452 Create the Partition Quota Configuration Files . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .479 In This Chapter . . .476 Limiting root Access . . . . . . . . . . . . . . . . . . .conf File .487 Configuring the NIS Client . . . . . . . . . . . . . . . . .473 Other NFS Considerations . . . . . . . . . . . . . .475 NFS Hanging . . . . . . . . . . . . .472 Troubleshooting NFS . . . . . . . . . . . . . . . . . . . . . . . . .480 Configuring the NFS Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .486 Adding New NIS Users . . . . . . . . . . .491 Configuring NIS Slave Servers . . . . . . . . . . . . . . . . . . . . . . . . . .468 The Structure of Direct and Indirect Map Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .485 Start the ypbind and ypxfrd Daemons . . . . . . . . . . . . . . . . . . . . . . . . . . . . .486 Make Sure the Daemons Are Running .480 Configuring the NFS Client . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .491 Configuring NIS Clients with Slaves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .488 Verify Name Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .479 Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .476 Conclusion . . . . . . . . . . . . . . . . . . .489 Test NIS Access to the NIS Server . . .483 Start the Key NIS Server-Related Daemons . . . . . . . . . . . . . . . . . . . . . . .490 NIS Slave Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .471 Starting automounter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .488 Start the NIS Client-Related Daemons . . . . . . . . . . . .476 Restricting Access to the NFS Server . . . . . . . . . . . . . . . . . . . . .472 The showmount Command . . . . . . . . . . . . . . . . .473 The df Command . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xxv Contents xxv The NFS automounter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .495 When There Is a NIS Master and Slave Pair . . .481 Configuring the NIS Server . . . . . . . . .468 automounter Map Files . . . . . . . . . .476 File Permissions . . . . . . . .496 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .475 File Locking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .495 Changing Your NIS Passwords . . . .469 Wildcards in Map Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .483 Edit Your /etc/sysconfig/network File . . . . . . . . . . . . . . . .477 30 Centralized Logins Using NIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .473 The nfsstat Command . . . . . . . . . . . . . . . .489 Test Logins via the NIS Server . . . . . . . . . . . . . . . . . . . . . . . .488 Run authconfig . . . . . . . . . . . .474 Security . . . . . . . . . . . . . . . . . . . . . . . . . . .471 automounter Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .483 Edit Your /etc/yp. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .496 Possible Password Errors . . . .476 Nesting Exports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .495 When There Is Only an NIS Master . .484 Initialize Your NIS Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .517 LDAP Web Management Tools . . . . . . . . . . . . . . . . . . .513 Possible stunnel Errors in Fedora Core 2 . . . . . . .qxd 2/25/05 12:39 PM Page xxvi xxvi Contents Considerations for a Non-NFS Environment . . . . . . . . . . . . . .502 Start the ldap Daemon . . .498 31 Centralized Logins Using LDAP and RADIUS . . . . . . . . . . . . . . . . . . . . . . . . . . .507 Configuring Encrypted LDAP Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .503 Modify the LDIF Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .515 Adding New LDAP Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .509 Troubleshooting LDAP Logins . . . . . . . . .499 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .501 Required LDAP Client RPMs . . . . . . . . . . . .512 Use the tcpdump Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .513 LDAP bind Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .499 The LDAP Directory Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .506 Edit the /etc/nsswitch File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .501 Configuring the LDAP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .512 Test Basic Connectivity . . . . . . . . . . . . . . . . . . .505 Edit the ldap. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .508 Configuring the stunnel LDAP Client . . . . . . . . . . . . . . . . . . . . . . . . .497 Conclusion . .501 Create a Database Directory . . . . . . . . . . . . . . . . . . . . .501 Create an LDAP Root Password . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .500 Scenario . . . . . . . . . . . .502 Convert the /etc/passwd File to LDIF Format .497 NIS Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .511 Test Using ldapsearch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .513 stunnel Doesn’t Appear to Work . . . . . . . . . . . . . . . . . . . . . . . . . . .513 LDAP Works But Is Not Using LDAPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .514 Modifying LDAP Users by User root . . . . . . . . . . . . .511 Use SSH or the Linux Console . . . . . . . . . . . . . . . . . . . . . . . . . . .516 Deleting LDAP Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf Configuration File . . . . . . . . . . . . . . . . . . . . . . . . .505 Test the LDAP Database . . . . . . . . . . . .500 Downloading and Installing the LDAP Packages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .508 Configuring the stunnel LDAP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .506 Testing . . . . . . .505 Configuring the LDAP Client . . . . . . . . . .517 .514 Starting and Stopping LDAP . . . . . . . . . . . . . . . .501 Required LDAP Server RPMs . . . . . . . . . . . . . . .504 Import the LDIF Files into the Database . . . . . . . . . . . . . . . .502 Edit the slapd. . . . . . . . . . . . . . . . .506 Create Home Directories on the LDAP Client . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .513 Common LDAP Administrative Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .514 LDAP Users Changing Their Own Passwords . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf File . . . . . . .542 Build the Kernel’s Modules . . . . . . . . . . . . . . .525 Restricting Web Access by Time . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . .530 Troubleshooting Squid . . . . . . . . . . . . . .536 Loading Kernel Modules on Demand . .540 Configure Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .537 Creating a Custom Kernel . . . . . . . . . . . . . . . . . . . . . . . . . . .520 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . .533 The Pros and Cons of Kernel Upgrades .524 The /etc/squid/squid. . . . . . . . . . . . . . . . . . . . . . . . . .520 Troubleshooting and Testing RADIUS . . . . . . . . . . . .config File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .523 Download and Install the Squid Package . . . . . . . . .523 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .517 Download and Install the FreeRADIUS Packages . . . . . . . . . . . . . . . . . . . .519 Configuring the /etc/raddb/clients. .526 Forcing Users to Use Your Squid Server . . .conf File . . . .config File .535 Reasons for Kernel Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .542 Compile a New Kernel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .530 Conclusion .539 Customizing the . . . . . . .525 Restricting Web Access by IP Address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .528 Making Your Squid Server Transparent to Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .534 Kernel Modules . . . . . . . . . . . . . . . . . . . . .535 How Kernel Modules Load When Booting . . . . . .542 Install the Kernel Modules . . . . . . . . . . . .531 33 Modifying the Linux Kernel to Improve Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .529 Squid Disk Usage . . . . . .qxd 2/25/05 12:39 PM Page xxvii Contents xxvii Configuring RADIUS for LDAP . . . . . .524 Starting Squid . . . . . . . . . .522 32 Controlling Web Access with Squid . . . . . . . . . . . . . . . . . . . .543 Copy the New Kernel to the /boot Partition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .524 Access Control Lists . . . . . . . .518 Starting and Stopping FreeRADIUS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .539 Make Sure Your Source Files Are in Order . . . . . . . . . . . . . . . . . . . .533 In This Chapter . . . . . . . . . . . . .534 Download and Install the Kernel Sources Package . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .528 Manually Configuring Web Browsers to Use Your Squid Server . . . . . . . . . . . .518 Configuring the /etc/raddb/radiusd. . . . . . . . . . . . . . . . . . . . . .539 The . . . . . . . . . . . .524 The Visible Host Name . . . . . . . . . . . .543 . . .542 Edit the Makefile to Give the Kernel a Unique Name . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .526 Password Authentication Using NCSA . . . . . . . . . . . . . . . . . .conf File .518 Configuring the /etc/raddb/users File . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .552 Viewing Your New MySQL Databases . . . . . . . .549 Creating a MySQL Root Account . . . . . . . . .553 Viewing Your MySQL Database’s Table Structure . . . . . . . . . . . . . . . .cnf File . . . . . . . . . . . . . . . .566 . . . . . . . . . . . . . . .565 Restart Openswan . . . .549 Accessing the MySQL Command Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .557 Conclusion . . . . . . . . . . . . . . . . . . . . .559 In This Chapter . . . . . .556 Basic MySQL Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .552 Listing the Data Tables in Your MySQL Database . . . . . . . . . . . . . . . . . . . .554 Recovering and Changing Your MySQL Root Password . . . . . . . . . . . .547 Preparing MySQL for Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .563 The /etc/ipsec. . . . . . . . . . . . .554 MySQL Database Backup . . . . . . . . . . . . . . . . . . . .545 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .560 Download and Install the Openswan Package . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .557 Test Database Access . . . . . . . . . . . . . . . . . . . . . . . . .550 Creating and Deleting MySQL Databases . . .547 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .548 Installing MySQL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .556 Very Basic MySQL Network Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .555 MySQL Database Restoration . . . . . . . . . . . . . . . . . . . . . . . . . .549 The Location of MySQL Databases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf Configuration File . . . . . . . . . . . . . . . .544 How to Create a Boot CD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .558 35 Configuring Linux VPNs . . . . . . . . . . .561 How to Get Openswan Started . . . . . . . . . .549 The /etc/my. . . . . . . . . . . . . .548 Starting MySQL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .560 Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .557 Connectivity Testing . . .557 A Common Fedora Core 1 MySQL Startup Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .551 Running MySQL Scripts to Create Data Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .550 Granting Privileges to Users . . . . . . . . . . . . . . . . . . . . . . . . . . .564 Edit the /etc/ipsec. . . . . . .546 34 Basic MySQL Configuration . . . . . . . . . . . . . . . . . .561 Get the Status of the Openswan Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .559 VPN Guidelines . .553 Viewing the Contents of a Table . . . . . . . . . . . . . . . . .conf File . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . .544 Kernel Crash Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . .554 Configuring Your Application . . . . . . . . . . . . . . . . . . . . . . . . . . .544 Updating the Kernel Using RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .562 How to Fix Common Status Errors .qxd 2/25/05 12:39 PM Page xxviii xxviii Contents Updating GRUB . . . . . .562 VPN Configuration Steps Using RSA Keys . . . . . . .563 Obtaining RSA Keys . . .

. . . . . . . . . . . . . . . .584 IKE’s Role in Creating Security Associations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .593 The mail-filter. . . . . . . . . . . . . . . . . . . . . . . . . . . .567 Possible Changes to IP Tables NAT/Masquerade Rules . .567 How to Ensure Openswan Starts When Rebooting .575 Linux Security with TCP Wrappers . . . .579 Configuration Steps . . . . . . . . . . . . .570 Check the Routes . . . . . . . . . . . . . . .571 Check syslog Error Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .593 The mail-filter Script . . . . . . . . . . . . . . . . . . .569 Testing VPN Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .576 The TCP Wrappers File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .572 Conclusion . . . . . . . . . . . . . . . . . . . .576 Determining the TCP Daemon’s Name . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .600 . . . . . . . . . . . . .588 Appendix II Codes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .581 Configuration Steps . . . . . . . . . . . . .569 Determine the Tunnel Status . . . . . . . . . . . . . . . . . . . . .599 NTP Server Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .593 iptables Scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .579 Preparing to Go Headless . . .586 VPN User Authentication Methods for Temporary Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .566 Testing the New Tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .598 FTP Client Script . .571 Using tcpdump . .600 Home/Small Office Protection Script . . . . . . . . . . . . . . . . . . . . . .572 Invalid Key Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .573 Appendix I Miscellaneous Linux Topics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .580 Make Your Linux Box Emulate a VT100 Dumb Terminal . . .581 VPN Terms and Methods . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .575 Fedora Core 3 . . . . . . . . . . . . . . . . . . . .585 VPN Security and Firewalls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .598 FTP Server Script . . . . . . . . . . . . . . . . . . . . .accept File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .587 ICMP Codes . . . . . . . . . . . . . . and Configurations . . . . . . . . . . . . . . .577 Running Linux Without a Monitor . . . . . . . . .591 Apache File Permissions Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .586 TCP/IP Packet Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .568 Using Pre-Shared Keys (PSK) . . . . . . . .qxd 2/25/05 12:39 PM Page xxix Contents xxix Initialize the New Tunnel . . . . . . . . . .577 Adjusting Kernel Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .568 Troubleshooting Openswan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .reject File . . . . . . .584 Internet Key Exchange (IKE) . . . . . . .591 sendmail Spam Filter Script . .583 Authentication and Encryption Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .592 The mail-filter. . . . . . . . . . . . . . . . . . . . . . . . . . Scripts. .

. . . . . . . . . . . . . . . . . . . .622 Appendix IV syslog Configuration and Cisco Devices . . . . . . . . .615 Appendix III Fedora Version Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .611 Zone File for my-web-site. . . . . . .613 Sample /etc/mail/access File . . . . . . . . . . . . . . . . .org (External View) . . . . . . . . .613 Sample /etc/mail/virtusertable File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .623 Cisco Local Director . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .609 The /etc/named. . . . . . . . . . . . . . . . . . . . . . . .619 iptables Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .618 Routing Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .619 Wireless Networking Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .627 . . . .613 Sample /etc/mail/sendmail. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .621 indexmaker MRTG_LIB Errors with Red Hat 9 and 8. . . . . . . . . . . . . . . . . . . . .623 Catalyst CAT Switches Running CATOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .617 FTP Differences . . . . . . . . .613 Sample /etc/mail/local-host-names File . . . . . . . . . . . . . . . . . . . .625 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .mc File . . . . . . . . . . . . . . . . . . . . . . .621 File Locations . . . . . . . . . . . . . . . . . . . . . . . . . . . . .0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .622 Webalizer Differences . . . . . . . . . . . . . . . . . . . . . . .625 The Sample Cisco syslog. . . . . . . . . . . . . . . . . . . . . . . . . .611 Reverse Zone File for a Home Network Using NAT . . . . . . . . . . . . . . . . . . . . . . .612 sendmail Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .624 Cisco PIX Firewalls . . . . . . . . . . . . . . . . . .620 Linux-WLAN File Locations Using Red Hat 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .621 Precedence Bitwise Error with Red Hat 9 . . . .609 Zone File for my-web-site.qxd 2/25/05 2:31 PM Page xxx xxx Contents Sample DNS Zone Files: Using BIND Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .org (Internal View) . . . . . . . . . . . . . . . . . . . . . . . . . .624 Cisco CSS11000 (Arrowpoints) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .619 Software Installation Differences . . . . . . . . . . . .618 DNS Differences . . . .conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .0 RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . .620 MRTG Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .617 DHCP Differences . . . . . . . .623 Cisco Routers . . .

Unfortunately. ® . and text editors. The notebook also includes many of the most commonly encountered errors with explanations of their causes and how to fix them. IT professionals frequently have to purchase additional specialist books on each topic. WHY IS THIS BOOK NECESSARY? Most Linux “encyclopedias” are split in three sections: an introductory section covering topics such as CD-based Linux installation. there are few practical texts in between that help you to make the transition from being a beginner to having the confidence of an expert.qxd 2/25/05 12:39 PM Page xxxi Preface Without question. The specialist guides are highly detailed. Like its many uses. Linux is rapidly becoming the operating system of choice in many core areas of business. It provides all the expected screen output when configuring the most commonly used Linux applications to help assure readers that they are doing the right thing. each with its own GUI interface. an intermediate section covering Microsoft Office clone productivity suites. from being used in products ranging from cell phones and PDAs to cars and mainframe computers. GUI interfaces.00Harrison_FM. but often are short on space to adequately cover detailed configuration instructions. Linux has a variety of printed and electronic guides to show you what to do. Unfortunately the “advanced sections” in these guides cover the underlying theory reasonably well. this guide exclusively uses the command line to illustrate the tasks needed to be done. focusing on narrow areas of excellence. Linux Quick Fix Notebook takes these “advanced sections” and expands them sufficiently to provide a practical tutorial guide on how to do basic configuration of many popular Linux back-office applications with command-bycommand instructions. It is transforming information technology in many exciting ways. and an advanced section covering the topics most nondesktop-support IT professionals use on a day-to-day basis. To avoid confusion between the many flavors of Linux. The encyclopedic guides for beginners focus on Linux fundamentals and only then introduce you to more specialized topics.

small-office. but as a trusted and experienced coworker. the persons most likely to be interested in this book would be IT professionals and consultants. APPROACH With this in mind. which provides room for much more coverage of the steps needed to get the more difficult jobs done.00Harrison_FM. computer literate business owners. and every chapter has real-world practical examples in the form of tutorials. The first starts as an introduction to networking and extends into using Linux as a main departmental server. Finally. students. only the essential supporting theory is provided to help end users implement their projects under budget and ahead of schedule. and DNS server setup Web sites on DHCP Internet links Time synchronization with NTP Error reporting with syslog Restricting users’ disk space usage with quotas . The chapters have a logical flow of information starting with concise backgrounders and ending with a troubleshooting section. Three sections have been created to make this process easier. A great deal of attention has been paid to troubleshooting techniques that are often needed to remedy unexpected behavior. manage.qxd 2/25/05 12:39 PM Page xxxii xxxii Preface PREREQUISITES The book’s format is aimed at proficient beginners. power users. and IT professionals who often have to do advanced tasks in which the underlying theory is understood. community college and trade school professors and students. They cover these essential topics: Software installation Networking setup and troubleshooting Samba for Windows files on Linux servers Linux-to-Linux file sharing with NFS Simple MySQL database administration LDAP and NIS for centralized logins FTP and SCP file transfers Disk drive redundancy with software RAID Wireless Linux networks MRTG server performance monitoring Linux firewalls and VPNs Squid for Web access control Mail. the third section covers more advanced topics that will become invaluable as your Linux administration role expands Explanations are given not as a lecturer. and monitor your own Linux-based Web site running on a simple DSL or cable modem line. Because the readers are assumed to be exposed to the theory of Linux. Additionally. Web. many of the introductory topics are not covered. basic Windows exposure to the concept of sharing directories between servers is needed. or home network and shows you how to set up the Linux servers most businesses need. but the commands to do it are forgotten or at the tips of their tongues. and SOHO workers. To maintain its appeal as a compact guide. Two to three months of hands-on Linux experience is an ideal prerequisite. The next section expands on this knowledge to show you how to create. The book creates a typical departmental.

. In summary. it becomes harder to remember and do it all. and homes ☞ Provides an excellent networking familiarization and troubleshooting guide As the line between power users and administrators continues to blur. as computers move from the data center to the desktop. this book ☞ Shows how to expand inexpensively an existing IT investment in Windows using Linux as the anchor of a network and Web site ☞ Shows how to do this command by command ☞ Is highly focused on being task-oriented ☞ Illustrates how to create a simple network for small business. and as Linux and Windows gain equal footing in business.qxd 2/25/05 12:39 PM Page xxxiii Preface xxxiii Many of the topics are covered in Linux certification exams making the book a valuable study guide for those seeking new areas of professional development.00Harrison_FM. corporate departments. This is the guide that will give you enough time to eat lunch.

Also includes a discussion on how to simply migrate a Web site from a third-party provider to your office or home.qxd 2/25/05 12:39 PM Page xxxiv xxxiv Preface BREAKDOWN PART 1 OF BOOK The Linux File Server Project Chapter 1 Why Host Your Own Site? A discussion on the pros and cons of hosting your own Web site. Chapter 5 Troubleshooting Linux with syslog Explanations on how to view and use Linux system error messages as part of the troubleshooting process. Chapter 8 Configuring the DHCP Server Essential topics on how to configure a DHCP server so that PCs are automatically assigned their Internet addresses when they boot up. Coverage of how to activate/deactivate the startup of various important system programs when the system boots up. Chapter 7 The Linux Boot Process Explanations of how Linux starts up after turning on the power switch. Chapter 4 Simple Network Troubleshooting Very detailed yet simply explained coverage of the various methods you can use to identify network trouble spots.00Harrison_FM. Details on how to choose between using the graphical and regular text interfaces on the Linux VGA console. . Chapter 3 Linux Networking A practical guide to configuring a network interface and simple Internet routing using Linux. Chapter 2 Introduction to Networking Covers the basic concepts needed to make the advanced user comfortable with many frequently encountered networking scenarios. Targets usage in the simple networks found in small offices and schools. Chapter 6 Installing RPM Software All you need to know on installing Linux software using popularly available RPM installation files. Describes 30 of the most commonly used terms with simple one-paragraph discussions.

and a sample iptables script is provided. This makes it easier to determine precisely who did what on a system. . Also provides explanations. corporate departments. There is also a tutorial on how to set up an FTP server to allow people to download only.qxd 2/25/05 12:39 PM Page xxxv Preface xxxv Chapter 9 Linux Users and sudo The basics on how to create Linux user accounts. Chapter 12 Samba Security and Troubleshooting Simple Samba problem resolution outlined in a logical step-by-step fashion. and Samba How to make your Linux server act transparently as a Windows file server using the Samba package. Chapter 13 Linux Wireless Networking Coverage of how to install. and test a wireless network that includes Linux servers. PART 2 The Linux Web Site Project Chapter 14 Linux Firewalls Using iptables Security is an essential part of Web life. files for the sake of security.00Harrison_FM. Linux. Chapter 11 Sharing Resources with Samba Explains how to share a Linux attached printer or CD-ROM drive with other Windows users. Chapter 15 Linux FTP Server Setup Shows how to set up a Linux server to allow uploading and downloading of files from the Internet using the File Transfer Protocol (FTP) package. This chapter gives an explanation of the most popular Linux-based firewall software with step-by-step examples of a script/program you can write to protect your network. with examples. of how to grant regular trusted users with temporary system administrator privileges and track what they do. configure. which is the typical scenario in schools. which can be virtually impossible to determine if everyone shares the administrator password. FTP can be tricky to configure with firewalls. not upload. Not all the features—just enough so that a user will be comfortable in creating one if software installation requires it. Uses a sample network of a few PCs. Chapter 10 Windows. and homes.

Also offers an example of how to save the configuration files of various Cisco networking devices to a Linux server using TFTP. and xinetd Coverage of how to log into a Linux server from a remote PC using the commonly available Telnet program. Chapter 18 Configuring DNS How to configure the Domain Name Services (DNS) on a Linux server that will enable you to map an Internet address to a Web site name. Chapter 21 Configuring Linux Mail Servers Covers how to configure the sendmail mail server to manage mail for your Web site. TFTP. but it can be easily simplified using macros that are provided with it.00Harrison_FM. . as are the steps needed to retrieve your mail from your sendmail server. Chapter 17 Secure Remote Logins and File Copying Telnet is very insecure as the data passing between the client and server is unencrypted. Chapter 20 The Apache Web Server After DNS is configured you need to set up a server to handle your Web site’s pages. The Apache Web server package is the most commonly used Linux software for this task. Chapter 19 Dynamic DNS In many cases you might want to host a Web site on a DSL line.qxd 2/25/05 12:39 PM Page xxxvi xxxvi Preface Chapter 16 Telnet. but the configuration file can be difficult to understand. Methods on how to combat SPAM. This chapter covers how to set up a secure shell (SSH) server for encrypted communications as an alternative to Telnet. Both the Telnet and TFTP packages are governed by the xinetd program. Many home and office networks get their Internet addresses dynamically assigned. sendmail can be difficult to configure. This chapter covers how to track this constant change so that your Web site’s name (URL) always points to your new Internet address. Explains the different types of DNS servers and discusses simple methods of how to make them serve your departmental network. Regular DNS relies on an Internet network whose Internet address doesn’t change. which is all explained in this chapter. Also covers how to configure SSH to allow you to log in to a remote Linux server without a password and to execute commands on a remote Linux server on demand. Full examples provided. which is also explained. one of the Internet’s greatest nuisances. are also explained. This chapter explains how to create a Web site simply by tacking on easy-tounderstand customizations to the end of the configuration file.

and disk usage statistics with examples. MRTG.00Harrison_FM. memory. VMSTAT. and even get Web-based Internet traffic graphs. Chapter 28 Managing Disk Usage with Quotas Examples of how to limit the amount of disk space individual Linux users can use on a filesystem. Chapter 29 Remote Disk Access with NFS A practical guide on how to use the network file system (NFS) to share files between Linux systems. Chapter 24 The NTP Server Explanations on how to synchronize the time on your Linux server with wellknown time servers on the Internet. Chapter 26 Linux Software RAID Coverage of how to configure Linux software to create redundant arrays of identical disk drives (RAID) to provide fail-safe data storage. . Also covers how to automate installation for large-scale deployments. using the SNMP. Chapter 30 Configuring NIS A description of how to use name information services (NIS) to create a centralized username/password database for all your Linux servers. with examples. Webalizer. TOP. PART 3 Advanced Topics Chapter 25 Network-Based Linux Installation How to install Linux quickly over a network without the need for CDs. which explains how to get graphs of CPU. as well as how to make your Linux server the primary time source for your network. Chapter 27 Expanding Disk Capacity Sometimes you run out of space on a Linux partition and need to expand it onto another hard disk. and FREE utilities.qxd 2/25/05 12:39 PM Page xxxvii Preface xxxvii Chapter 22 Monitoring Server Performance Very detailed coverage on how to monitor your server’s performance. This chapter explains how to do it. Chapter 23 Advanced MRTG for Linux Expanded coverage of MRTG.

Also covers how to provide password-only access to the Internet. and Configurations Provides the library of useful sample configuration files and productivity enhancing scripts that are used as references in the main text. Chapter 33 Modifying the Kernel to Improve Performance A brief explanation with examples on how to configure the Linux kernel to improve system performance. including additional security using TCP wrappers. Chapter 34 Basic MySQL Configuration Many add-on Linux applications require interaction with a database of some sort.00Harrison_FM. Appendix IV syslog Configuration and Cisco Devices Shows how to configure many popular Cisco network devices to send their status and error messages to a Linux server. This is a handy brief reference when you are considering an upgrade of your operating system. it can be used to authenticate users logging into a variety of devices. Scripts. Appendix III Differences Between The Fedora Versions Describes many of the subtle differences in the functioning and configuration of the various versions of Fedora and Red Hat Linux.qxd 2/25/05 12:39 PM Page xxxviii xxxviii Preface Chapter 31 Centralized Logins Using LDAP and RADIUS NIS has its limitations because it is geared toward Linux logins and does unencrypted authentication. adjusting dynamic kernel parameters. This chapter discusses basic MySQL configuration for the purposes of supporting the installation of third-party Linux applications. This was included because many Linux administrators also have a networking and security management role as well. a description of Security Enhanced Linux. The lightweight directory access protocol (LDAP) can also be used to store username/password data in a rapid access database format over an encrypted network connection. Chapter 35 Configuring Linux VPNs Details of how to create a permanent site-to-site VPN between two offices. . Chapter 32 Controlling Web Access with Squid This chapter focuses on how to use Squid to limit the amount of time users can spend browsing the Internet using the Squid Linux package. When used in conjunction with RADIUS. topics for Linux servers. such as accounting packages and message boards. yet often overlooked. one of the most popular is MySQL. and how to manage your Linux server via a connection to the COM port. such as networking equipment. Appendix II Codes. This chapter explains how. Appendix I Miscellaneous Linux Topics Covers a variety of useful. that don’t run Linux.

00Harrison_FM.qxd 2/25/05 12:39 PM Page xxxix .

Thank you all. and George Vish. One of them. correcting errors. Also. Bret Strong. Jack. too many to completely acknowledge on a single page. Barbara. Robert Husted. visitors to the site have sent numerous e-mails of encouragement. Blaze. Jade. provided me with the extra push to seek publication. without the combined knowledge of my coworkers the task would have been nearly impossible. I must first thank my family—Diana.qxd 2/25/05 12:39 PM Page xl Acknowledgments There have been numerous people who have provided me with the inspiration to write. who challenged me to improve the technical content and readability of the book. Christy Hackerd. . who had the vision to take a chance. Ron McCarty. It would also be unthinkable not to mention Jill Harry of Prentice Hall. Ever since that simple start. and Linda Laflamme who always directed me along the right path.00Harrison_FM. with Brenda Mulligan. Keith Lofstrom. and Helene—who provided encouragement since the book was a single Web page on wireless Linux. suggesting new directions. Nancy Albright. Pauline. But the real dessert was reading the reviews of (alphabetically listed) Joseph Brazeal. Without them all the cake would have no icing. and finding omissions. Joseph Naberhaus.

00Harrison_FM.qxd 2/25/05 12:39 PM Page xli About the Author Peter Harrison has been fascinated by computers since the early 1980s. Peter likes to relax with his family on short weekend trips to the many attractions of the San Francisco Bay Area. the techie. Extensive use of Linux in this environment combined with his varied business background has helped him create this highly readable book for the newbie. In his quieter moments. and their bosses. and was the principal systems engineer responsible for the computerization of the island’s tax collection and social security systems. Jamaica’s first computer user group. Long rides on his bicycle provide another guilty pleasure. He was the founding president of PCJAM. Peter enjoys the art and literature of the Caribbean and Latin America. He then sought new opportunities as the western Caribbean representative for a Fortune 500 pharmaceuticals firm and later became the international sales manager for a West Indian rum company. Peter has since worked extensively in the Internet sector deploying large-scale data centers and Web sites. . Before moving to Silicon Valley he ran Trinidad and Tobago’s first industrial trade office to Latin America.

00Harrison_FM.qxd 2/25/05 12:39 PM Page xlii .

01Harrison_Pt1.qxd 2/25/05 9:47 AM Page 1 P A R T I The Linux File Server Project 1 .

qxd 2/25/05 9:47 AM Page 2 .01Harrison_Pt1.

qxd 2/25/05 9:46 AM Page 3 C H A P T E R 1 Why Host Your Own Site? IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ Our Network Alternatives to In-House Web Hosting Factors to Consider Before Hosting Yourself How to Migrate From an External Provider In-House Server Considerations Conclusion Web sites have proliferated greatly over the years to become a part of everyday life for many people. This chapter briefly addresses the most common issues and outlines the simple network architecture for use in a small office or home on which the focus of the rest of the book will be based. but later expanded them to become an important part of their operations. Many companies rely almost exclusively on their Web sites to sell their products and provide both customer and supplier support services. With this in mind. People use them to create Web logs of their daily lives. the first 3 .linuxhomenetworking. the book is divided into three sections of gradually increasing complexity to make this process easier. You have to consider factors of cost and convenience as well as service and support.02Harrison_ch01.com. but the process of establishing one touches many aspects of not only Linux. The following is a typical Web site address: www. Businesses originally used Web sites primarily as a marketing tool. Not all businesses departments and homes require a Web site. provide family members with a place to store their memories. This book is about setting up Linux servers to do the things that most businesses and homes need. After this chapter. but information technology as well. or tell people of their experiences in getting things to work. It’s about getting the job done. The decision as to whether to host your own Web site can be difficult.

253. Finally.1 Wireless home network topology. As stated before.25 (Depends on Tutorial) Wireless Card Internet Internet User ISP Gateway Firewall/Router WAP Linux Box (BIG BOY) 192. manage. but before you do it would be best to weigh the pros and cons of doing this.1. and troubleshooting before the first major project of using Linux as a main departmental file server for Windows PCs. the third section covers more advanced topics that will become invaluable as your Linux administration role expands. as seen in Figure 1.30 Outside IP Address: DHCP or 97.158. the rest of the book shows how you can make a simple layout such as this become a functional low-volume Web site. ALTERNATIVES TO IN-HOUSE WEB HOSTING There are two broad categories of hosting options for small Web sites. There are also those that allow you to use servers completely dedicated to your site.1. with a router/firewall connected to a broadband Internet connection protecting a single network on which all servers and PCs are connected.02Harrison_ch01.100 Inside IP Address: 192.1.253.158. The next section expands upon this knowledge to show you how to create. and monitor your own Linux-based Web site on this network using a simple DSL or cable modem Internet connection. Wireless Home Network Topology Public Network Private Network ISP Gateway's IP Address: 97. OUR NETWORK The typical small office or home network is usually quite simple. software installation.qxd 2/25/05 9:46 AM Page 4 4 Why Host Your Own Site? Chapter 1 section introduces you to networking.168. these are called dedicated hosting . There are companies that host multiple Web sites on the same server and are called virtual hosting providers.168.1 Wireless Card Workstation DHCP Figure 1.

from which to choose. 2.my-web-site. you typically have to make contact with a live sales representative who represents an Internet data center. 3. and the name of a private directory on a shared Web server in which you’ll place your Web pages. which allows you to have a shopping cart and customer loyalty programs. a number of e-mail addresses and an easy-to-use Web-based GUI to manage your settings. The disadvantage of virtual hosting is that. and the amount of Internet bandwidth you expect to use. These may not necessarily be to your liking. Start testing your site using your IP address in your Web browser. It takes about 3-4 days for DNS to propagate across the Web. Sign up for the virtual hosting service. you often have no control over the operation of the server and have to rely on the staff and operational procedures of the hosting company to get your changes implemented.com. Upload your Web pages to your private virtual hosting directory. www. so you’ll probably have to wait at least that long before you’ll be able to view your site using your sub-domain. and bandwidth graphing information often can be purchased as extra line items on your bill. many will also provide an e-commerce feature. such as my-web-site. 4. or RegisterFree.qxd 2/25/05 9:46 AM Page 5 Alternatives to In-House Web Hosting 5 providers. As you can imagine. A 1 1⁄2-inch slot in a computer cabinet for a Web server can easily cost $200 per month for .02Harrison_ch01.com. Verisign. firewall management. They will provide you with a user account.org. the IP address of your site. The steps are fairly straightforward: 1. The virtual hosting provider will also offer free backups of your site. Register your domain name. Dedicated providers might provide you with only a network connection for a server you purchase and install in their data center. At a minimum. For an additional charge. on-call engineering time. technical support. Dedicated Hosting In this scenario. Virtual Hosting It is easy to find virtual hosting companies on the Web that offer to host a simple Web site for about $10 per month. with companies like Register. these services can be fairly expensive. from monitoring to backups. the amount of power it consumes. though it is cheap. you have to pay for the amount of space your server occupies in the data center. monitoring. You must make sure your new domain name’s DNS records point to the DNS server of the virtual hosting company. or they might offer a menu of services.org. Additional services such as backups.

especially virtual hosting.1 Factors to Consider When Hosting In-House Costs New hardware and software Risks Likelihood of a failure and its expected duration The cost of both the failure and post-failure recovery (hardware. You can install the software you need. is the ideal solution for many small Web sites but there are a number of reasons why you may want to move your Web site to your home or small office. not what the ISP dictates.1. Table 1. data restoration. . you may also want to consider doing it yourself. software. Despite the relative merits of external hosting. The pros and cons of doing this are listed in Table 1. not those of the hosting provider. FACTORS TO CONSIDER BEFORE HOSTING YOURSELF Hosting your Web site externally. Pros ☞ Increased Control and Flexibility: You will be able to manage all aspects of your Web site if it is hosted on a server based either in-house or within your control at a remote data center. The advantage over virtual hosting is that you can customize the server to your needs. You won’t have to wait before changes are made and you can select the IT solution that best meets your needs. time) Irregular procedures that could increase the vulnerability of your site to failure Savings Monthly outsourced Web hosting fee Elimination of the cost of delays to implement desired services Possible new application development Training The percentage of IT staff’s time installing and maintaining the site Potential cost of the risks (% likelihood of failure per month × cost of failure) Is In-House Hosting Preferred? There are a number of advantages and disadvantages to hosting Web sites that are physically under your own control.02Harrison_ch01.qxd 2/25/05 9:46 AM Page 6 6 Why Host Your Own Site? Chapter 1 1Mbps of bandwidth.

You may not have access to cheap real estate in which to host your servers. redundant hardware. Availability: Reliable virtual hosting facilities may not be available in your country. You may also have to incur additional training costs to ensure that your IT staff has the necessary knowledge to do the job internally. engineering. Commercial office space is often more expen- . and customer care services. Cost: It is possible to host a Web site on most DSL connections.02Harrison_ch01. You will eventually be able to justify hosting your Web site in-house based on this financial fact. security audits. Hosting providers may provide software patches to fix security vulnerabilities on your Web servers and may even provide a firewall to protect it. load balancing. So for a savings of $10 per month the project should pay for itself in less than a year. DNS. Cons ☞ Lost Services: You lose the convenience of many of the services—such as backups. The cost of using an external Web hosting provider will increase as you purchase more systems administration services. an in-house solution may be better. New Skills: An additional benefit is learning the new skills required to set up the site. If you can’t get adequate support for billing. A Web site can be hosted on this data circuit for only the additional hardware cost of a network switch and a Web server. database services. there are other costs to consider. Proceed with the server migration only if you feel your staff can handle the job. ☞ Technical Ability: Your service provider may have more expertise in setting up your site than you do. and/or you may not have access to the foreign currency to host your site abroad. and technical support—offered by the virtual hosting company. ☞ Security: One important factor to consider is the security of your new server. These services may be more difficult to implement in-house. ☞ Availability: In many cases the reliability of a data center’s Internet connectivity is usually higher than that of your broadband connection. Language: ISPs often provide technical support in only a few languages. You should be able to buy this equipment secondhand for about $100. so that you can become a small virtual hosting service yourself. Always weigh the degree of security maintained by your hosting provider against the security you expect to provide in-house. If your home already has DSL there would be no additional network connectivity costs. Changes can be made with little delay.qxd 2/25/05 9:46 AM Page 7 Factors to Consider Before Hosting Yourself 7 ☞ ☞ ☞ ☞ There is also the possibility of offsetting the cost of your server by subleasing space on it to other companies in your community. ☞ Cost: Though you may be able to save money on a data circuit.

I have one server . Purpose-built Web servers tend to use multiple CPUs. receives more than half a million page views per month and runs on a 1GHz Intel Celeron with 1GB of RAM. special error-correcting ECC RAM. and additional power feeds. a basic desktop system will work fine. but you pay for the peace of mind when your only source of income is your Web site. You can place your small wireless router near your DSL/cable modem and the server anywhere in the house. high-speed redundant SCSI disks that can be replaced while the system is running without affecting performance. such as UPSs. Another good idea is to color-code your cables. has a detailed explanation of the steps involved in migrating your Web site from an external hosting provider to your home or small office. maintenance contracts.” and 21 “Configuring Linux Mail Servers.02Harrison_ch01.com. In my little lab. www. Wireless technology for a home-based Web site can be extremely convenient. all of which may be already bundled in with the services of an external data center.” on Web and mail server configuration to help provide a more rounded understanding of the steps involved.” which covers DNS. You should also read Chapters 20. A secondhand PC is adequate in this case. special built-in diagnostic tools. Some companies use one color for networks using private IP addresses and another for Internet-facing networks.qxd 2/25/05 9:46 AM Page 8 8 Why Host Your Own Site? Chapter 1 sive than basic data center space.linuxhomenetworking. backup systems. others use one color for straight-through and another for crossover cables. Try to have a dedicated area for your server that’s clean. and dry. dual redundant power supplies. monitoring systems. cool. “Configuring DNS. software patch management. IN-HOUSE SERVER CONSIDERATIONS For small Web sites without a great deal of database activity and where “hot standby” hardware isn’t a great need. They cost significantly more. and that uses UPS-protected power. which was the inspiration for this book. “The Apache Web Server. You may have to purchase additional equipment and services to support your servers. multiple PCI buses. Label all your cables at both ends and try to create an updated network diagram that you can show anyone who will provide you assistance. HOW TO MIGRATE FROM AN EXTERNAL PROVIDER Chapter 18. and slim-line cases only a few inches high.

in that the incoming downstream data rate from the Internet is different than the reverse outgoing upstream speed. You should be most concerned about upstream speed for your Web site to the Internet. These issues are becoming less important as software vendors are increasingly porting their applications to Linux. Red Hat Linux was a free product. Inbound Web browser queries don’t use a lot of data bandwidth. What’s the difference? Until Version 9. Selecting an Internet connection for your Web site is not as easy as it seems. They reserve symmetrical data circuits for businesses that usually need high bandwidth to both surf the Web and serve Web pages. its popular corporate cousin. there may be no other choice. one under a bed. New versions of Fedora are released every six months. Some technologies. an increasingly strong rival to Windows.qxd 2/25/05 9:46 AM Page 9 In-House Server Considerations 9 behind a bookcase. especially for smaller companies. such as T1 links. residential customers usually get a dynamic address allocation. Another source of concern would be deciding on the operating system to use. which may be the only product your Web or business application will work with and with which your staff is most familiar. A popular one is Windows. When you live in an apartment. the Fedora versions are maintained by Red Hat. another behind the TV. High-speed links are usually marketed to businesses and their cost per megabyte of data transfer is usually higher because the service may be combined with data center space. and these maintained the Red Hat brand. and wireless links.02Harrison_ch01. DSL and wireless services are sometimes asymetrical. but the Web pages that contain the outbound replies do. At the same time Red Hat decided to create Fedora Linux as a support-free product with an aggressive development cycle. The ISP will usually provide the business with a fixed range of Internet addresses as part of the service. but the service also has a per kilometer monthly distance charge. This can have an impact on your Web site preparation and will be discussed in more detail in later chapters. which also has a lower overall total cost of ownership. offer more bandwidth. DSL. The company then decided to create enterprise and desktop versions that had paid service contracts bundled with them. Internet service providers (ISPs) provide asymmetric services for residential users and the downstream rate is almost always higher than the upstream. and a couple around my desk. which is generally unsuitable for businesses that often require more stability and support. Once Fedora updates are proven . but they may not be available in your area or the installation times may not be acceptable. Though the original applications may be developed by volunteers. can optionally provide a dedicated circuit between two locations external to the Internet. be more reliable. There are many data circuit technologies such as cable modem. but the risk is that a book falling behind a bookcase or a bounce from a vacuum cleaner could take your site off the air. This book focuses on Fedora Linux with some references to Red Hat Linux. and provide better customer support.

. which are updated every 12–18 months. If you decide to do it. they are incorporated into the Red Hat Linux releases. Whatever you choose to do. this book will provide a lot of guidance in completing a successful project. and always be aware of the potential risks of the decision you make. I also chose Fedora Linux because it’s popular and it’s the Linux flavor I’ve worked with most frequently at home and at work.qxd 2/25/05 9:46 AM Page 10 10 Why Host Your Own Site? Chapter 1 stable. Always get a professional opinion. I chose Fedora because it’s free. You don’t have to get a purchase order to play with Fedora. and management. even if it’s informal.02Harrison_ch01. and other Linux distributions should also be considered. When you become comfortable with it and have proven the concept to yourself. This may not be the one suitable for you. plan carefully. CONCLUSION The decision to manage your Web site in-house can be difficult. you can then consider the more stable Red Hat equivalent. Constant communication between Red Hat and the developers help to keep the updates synchronized. your peers.

These chapters will then introduce the remaining chapters that cover Linux troubleshooting. therefore. Almost all computers are now networked in some way to other devices. and the configuration of many of the most popular Linux applications used in corporate departments and Small Office/Home Office (SOHO) environments. friends. It will help make the road to Linux mastery less perilous. general software installation.03Harrison_ch02.qxd 2/25/05 9:55 AM Page 11 C H A P T E R 2 Introduction to Networking IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ The OSI Networking Model An Introduction to TCP/IP How IP Addresses Are Used to Access Network Devices How Subnet Masks Group IP Addresses Into Networks The Physical and Link Layers Networking Equipment Terminology Additional Introductory Topics Conclusion Installing the Linux operating system is only the first step toward creating a fully functional departmental server or Web site. 11 . a basic understanding of networking and issues related to the topic will be essential to feeling comfortable with Linux servers. and even you. Familiarity with the concepts explained in the following sections will help answer many of the daily questions often posed by coworkers. This introductory chapter forms the foundation on which the following network configuration and troubleshooting chapters will be built. a road that begins with an understanding of the OSI networking model and TCP/IP.

When the data finally arrives. A good analogy would be a traveler who prepares herself to return home through many dangerous kingdoms by obtaining permits to enter each country at the very beginning of the trip. This layer also strips away its permit and forwards the data to the next layer. Converts data from one presentation format to another—for example. layer 3 ensures that the data can hop from link to link on the way to the final destination described in its header. The very first layer of the OSI model describes the transmission attributes of the cabling or wireless frequencies used at each link or step along the way. Application telnet FTP sendmail 5 Session .qxd 2/25/05 9:55 AM Page 12 12 Introduction to Networking Chapter 2 THE OSI NETWORKING MODEL The Open System Interconnection (OSI) model. she asks the border guards for directions to reach the next frontier and displays the permit for that new kingdom as proof that she has a legitimate reason for wanting to go there. Layer 2 describes the error correction methodologies to be used on the link. developed by the International Organization for Standardization.03Harrison_ch02. Each layer extracts the permit. defines how the various hardware and software components involved in data communication should interact with each other. a kingdom over which it rules. or header information.1. At each frontier. and so the cycle continues for seven layers. our friend has to hand over a permit to enter the country. A description of the functions of each layer in the model can be seen in Table 2. Once inside. e-mail text entered into Outlook Express being converted into SMTP mail formatted data. graphical on a screen or otherwise. In the OSI model each component along the data communications path is assigned a layer of responsibility—in other words.1 Layer 7 6 The Seven OSI Layers Name Application Presentation Description The user interface to the application. Finally. it needs from the data and uses this information to correctly forward what’s left to the next layer. The application uses the guidelines of layer 5 to keep track of the various communications sessions it has with remote computers and uses layer 6 to verify that the communication or file format is correct. layer 7 defines what the end user will see in the form of an interface. Table 2. the layer 4 header is used to determine which locally installed software application should receive it. Manages continuing requests and responses between the applications at both ends over the various established connections.

It is part of the larger OSI model upon which most data communications is based. Correctly resequences data packets that arrive in the wrong order.03Harrison_ch02. the resulting data is said to be a segment. the data is usually split into multiple pieces or packets. . which is responsible for ensuring that data is transferred between two addresses without being corrupted. For manageability. One component of TCP/IP is the Internet Protocol (IP). Defines the electrical and physical characteristics of the network cabling and interfacing hardware. Data sent on a link is said to be structured in frames. each with its own error detection bytes in the control section or header of the packet. the resulting data is said to be a packet. Handles the routing of data between links that are not physically connected.qxd 2/25/05 9:55 AM Page 13 An Introduction to TCP/IP 13 Layer 4 Name Transport 3 Network 2 Link 1 Physical Description Manages the establishment and tearing down of a connection. Ensures that unacknowledged data is retransmitted. After the packet’s overhead bytes have been stripped away. Error control and timing of bits speeding down the wire between two directly connected devices. Application TCP UDP IP ARP Ethernet ARP Ethernet AN INTRODUCTION TO TCP/IP TCP/IP is a universal standard suite of protocols used to provide connectivity between networked devices. After the link’s overhead bytes have been stripped away. It then passes the data to the program that expects to receive it. The remote computer then receives the packets and reassembles the data and checks for errors.

which is used to determine which network application on the computer should process the data. TCP Is a Connection-Oriented Protocol TCP opens up a virtual connection between the client and server programs running on separate computers so that multiple and/or sporadic streams of data can be sent over an indefinite period of time between them. This SYN.” The conversation ends with a final “bye” from the hungry person.” The communication then continues with a series of segment exchanges.03Harrison_ch02. it sends a segment to the other with the FIN and ACK bits set. TCP is a good example of a connection-oriented protocol. “Who is it?”. How TCP Establishes a Connection Any form of communication requires some form of acknowledgment for it to become meaningful. Programs that use TCP therefore have a means of detecting connection failures and requesting the retransmission of missing packets. Both persons knew who was on the other side of the door before it opened and a conversation can now begin. This is explained in more detail later. The two most popular transportation mechanisms used on the Internet are Transmission Control Protocol (TCP) and User Datagram Protocol (UDP). the TCP/UDP header is then inspected for the port value. to which the other server also replies with a FIN-ACK segment. TCP acts in a similar way. The server initiating the connection sends a segment with the SYN bit set in TCP header. to which the originating server replies with a segment with only the ACK bit set. This informs the computer receiving the data about the type of layer 4 transportation mechanism being used. to which the visitor replies. When one of the servers needs to end the communication. ACK mechanism is often called the “three-way handshake. “It’s me!” Then the door opens. Someone knocks on the door to a house. When the type of transport protocol has been determined. the person inside asks. . The communication terminates with a final ACK from the server that wanted to end the session. SYN-ACK. TCP keeps track of the packets sent by giving each one a sequence number with the remote server sending back acknowledgment packets confirming correct delivery. I have to go for lunch. The target replies with a segment with the SYN and ACK bits set.” to which the reply is. see you tomorrow. each with only the ACK bit set. “I think I’m finished here too.qxd 2/25/05 9:55 AM Page 14 14 Introduction to Networking Chapter 2 How does the computer know what program needs the data? Each IP packet also contains a piece of information in its header called the type field. This is the equivalent of ending a conversation by saying “I really have to go now.

This makes the second byte in the segment number 9767. ACK] Seq=8404 Ack=9767 Win=5792 -> -> -> -> -> -> hostB hostB hostA hostA hostB hostA TCP 1443 > http [ACK] HTTP HEAD / HTTP/1. Only the TCP segment header contains sequence information. The acknowledgment number. UDP. the third number 9768.03Harrison_ch02. a random value of 9766 was assigned to the first byte. TCP and UDP Ports The data portion of the IP packet contains a TCP or UDP segment sandwiched inside. If data isn’t received correctly. but both the UDP and the TCP segment headers track the port being used. the sequence number represents the serial number of the first byte of data in the segment. not to be confused with the ACK bit. The . or Ack. and all subsequent bytes for the connection from this host will be sequentially tracked.” You can clearly see the three-way handshake to connect and disconnect the session: hostA hostB Len=0 hostA hostA hostB hostB hostA hostB Len=0 hostA Len=0 hostB -> hostB TCP 1443 > http [SYN] Seq=9766 Ack=0 Win=5840 Len=0 -> hostA TCP http > 1443 [SYN.1 200 OK TCP 1443 > http [ACK] TCP http > 1443 [FIN.qxd 2/25/05 9:55 AM Page 15 An Introduction to TCP/IP 15 Here is a modified packet trace obtained from the tethereal program discussed in Chapter 4. the receiver will re-send the requesting segment asking for the information to be sent again. It is also used when data needs to be broadcast to all available servers on a locally attached network where the creation of dozens of TCP connections for a short burst of data is considered resource-hungry. and the total number of bytes cannot exceed the Win or window value that follows it. “Simple Network Troubleshooting. TCP’s “Connectionless” Cousin UDP is a connectionless protocol. The TCP code keeps track of all this along with the source and destination ports and IP addresses to ensure that each unique connection is serviced correctly.1 TCP http > 1443 [ACK] HTTP HTTP/1. So in the first line. is the sequential serial number of the first byte of the next segment it expects to receive from the other end. etc. Data is sent on a “best effort” basis with the machine that sends the data having no means of verifying whether the data was correctly received by the remote machine. UDP is usually used for applications in which the data sent is not mission-critical. Seq=9767 Ack=8405 Win=5840 Len=0 Seq=8405 Ack=9985 Win=54 Len=0 Seq=9985 Ack=8672 Win=6432 Len=0 ACK] Seq=8672 Ack=9985 Win=54 -> hostB TCP 1443 > http [FIN. ACK] Seq=9985 Ack=8673 Win=6432 -> hostA TCP http > 1443 [ACK] Seq=8673 Ack=9986 Win=54 In this trace.

The server sending the packet sets the initial TTL value. For example. and port 25 is reserved for SMTP e-mail. the network device will discard the packet. Remember this concept because it will be helpful in understanding the traceroute troubleshooting technique outlined in Chapter 4 that covers network troubleshooting. Usually when a connection is made from a client computer requesting data to the server that contains the data: ☞ The client selects a random previously unused source port greater than 1024 and queries the server on the destination port specific to the application. control. If the TTL value reaches 0. it tells the TCP application to respond back to port 2049 of the client using a source port of port 80. and informational messages for use by the operating system. ICMP provides a suite of error. When the Web server software replies to the client. If it is an HTTP request. The ICMP Protocol and Its Relationship to TCP/IP There is another commonly used protocol called the Internet Control Message Protocol (ICMP). Ports less than or equal to 1024 are reserved for privileged system functions. IP packets will occasionally arrive . but TCP/IP-based applications use it frequently. The TCP/IP Time to Live Feature Each IP packet has a Time to Live (TTL) section that keeps track of the number of network devices the packet has passed through to reach its destination. 2049 and query the server on port 80 (HTTP). ☞ The server recognizes the port 80 request as an HTTP request and passes on the data to be handled by the Web server software.qxd 2/25/05 9:55 AM Page 16 16 Introduction to Networking Chapter 2 source/destination port and the source/destination IP addresses of the client and server computers are then combined to uniquely identify each data flow. and those above 1024 are generally reserved for non-system third-party applications. Certain programs are assigned specific ports that are internationally recognized. and each network device that the packet passes through then reduces this value by 1. port 80 is reserved for HTTP Web traffic. ☞ The client keeps track of all its requests to the server’s IP address and will recognize that the reply on port 2049 isn’t a request initiation for “NFS. say.” but a response to the initial port 80 HTTP query. TTLs therefore help reduce the clogging of data circuits with unnecessary traffic. This mechanism helps to ensure that bad routing on the Internet won’t cause packets to aimlessly loop around the network without being removed. the client will use a source port of. It is not strictly a TCP/IP protocol.03Harrison_ch02. For example.

. ICMP TTL expired messages are also sent by network devices back to the originating server whenever the TTL in a packet is decremented to 0.0 .0.255 192.25.” which covers Linux specific networking topics. HOW IP ADDRESSES ARE USED TO ACCESS NETWORK DEVICES All TCP/IP-enabled devices connected to the Internet have an Internet Protocol (IP) address.” and Chapter 4.31. The server will usually detect this by examining the packet and correlating the contents to what it finds in the IP header’s error-control section.10.255. For ease of use. An example of an IP address that follows these rules is 97.65. Just like a telephone number.168. “Linux Networking. electrical interference. Note Chapter 3. explains how to configure the IP address of your Linux box. The Internet Assigned Numbers Authority (IANA) is the organization responsible for assigning IP addresses to Internet Service Providers (ISPs) and deciding which ones should be used for the public Internet and which ones should be used on private networks. ICMP also includes echo and echo reply messages used by the Linux ping command to confirm network connectivity.1.168.0 . each representing a number from 0 to 255. “Codes. More information on ICMP messages can be found in both Appendix II. These are called private IP addresses and have the following ranges: 10.1 to 192.255 172.0.255.qxd 2/25/05 9:55 AM Page 17 How IP Addresses Are Used to Access Network Devices 17 at a server with corrupted data due to any number of reasons including a bad connection. and Configurations.1.12.16.0.168.168.0. IP addresses are in reality a string of 32 binary digits or bits. You may be wondering how devices using private addresses could ever access the Internet if the use of private addresses on the Internet is illegal.255.255.03Harrison_ch02. network engineers often divide these 32 bits into four sets of 8 bits (or octets). or even misconfiguration.192.255.255 Home networking equipment/devices usually are configured in the factory with an IP address in the range 192. Scripts. It will then issue an ICMP reject message to the original sending machine saying that the data should be re-sent because the original transmission was corrupted.) to create the familiar dotted decimal notation.0 . Each number is then separated by a period (. Private IP Addresses Some groups of IP addresses are reserved for use only in private networks and are not routed over the Internet.172. it helps uniquely identify a user of the system.

This concept is important to understand. Don’t worry. Some of the terms mentioned here might be unfamiliar to you. You can configure NAT to be one to one. This is called network address translation (NAT) and is often also called IP masquerading in the Linux world. Network Address Translation (NAT) Makes Private IPs Public Your router/firewall will frequently be configured to give the impression to other devices on the Internet that all the servers on your home/office network have a valid public IP address and not a private IP address. the two most commonly stated are: ☞ No one on the Internet knows your true IP address. There are many good reasons for this. This greatly increases the number of devices that can access the Internet without running out of public IP addresses. As a general rule. .0. in which you request your ISP to assign you a number of public IP addresses to be used by the Internetfacing interface of your firewall. and then you pair each of these addresses to a corresponding server on your protected private IP network.” that covers the Linux iptables firewall. This prevents hackers from directly attacking your home systems because packets sent to the private IP will never pass over the Internet. this problem is overcome by NAT. in which the firewall maps a single IP address to multiple servers on the network.03Harrison_ch02. Basic NAT testing requires you to ask a friend to try to connect to your home network from the Internet. you won’t be able to access the public NAT IP addresses from servers on your home network. NAT protects your home PCs by assigning them IP addresses from private IP address space that cannot be routed over the Internet. This IP address is defined as 127.1 and is frequently referred to as localhost. This must cause networking confusion. and will be revisited in many later chapters. The localhost IP Address Whether or not your computer has a network interface card it will have a built-in IP address with which network-aware applications can communicate with one another.qxd 2/25/05 9:55 AM Page 18 18 Introduction to Networking Chapter 2 The situation gets even more confusing if you consider the fact that hundreds of thousands of office and home networks use these same addresses. but they will be explained in later sections of this chapter. ☞ Hundreds of PCs and servers behind a NAT device can masquerade as a single public IP address. “Linux Firewalls Using iptables. You can also use many to one NAT.0. Examples of NAT may be found in the IP masquerade section of Chapter 14.

DHCP The Dynamic Host Configuration Protocol (DHCP) is a protocol that automates the assignment of IP addresses.qxd 2/25/05 9:55 AM Page 19 How IP Addresses Are Used to Access Network Devices 19 Port Forwarding with NAT Facilitates Home-Based Web Sites In a simple home network. The reverse isn’t true. at which point the DHCP client and server renegotiate a new IP address from the server’s predefined pool of addresses. subnet mask. DNS server. New connections initiated from the Internet to the public IP address of the router/firewall face a problem. In this case the data is usually discarded. and other IP parameters. subnet mask default routers. Because the router/firewall is located at the border crossing to the Internet. . all servers accessing the Internet will appear to have the single public IP address of the router/firewall because of many to one NAT. The DHCP client sends out a query requesting a response from a DHCP server on the locally attached network. For example.03Harrison_ch02. it can easily keep track of all the various outbound connections to the Internet by: ☞ Monitoring the IP addresses and TCP ports used by each home-based server ☞ Mapping it to the TCP ports and IP addresses of the Internet servers with which they want to communicate This arrangement works well with a single NAT IP trying to initiate connections to many Internet addresses. you can configure your router/firewall to forward TCP port 80 (Web/HTTP) traffic destined to the outside NAT IP to be automatically relayed to a specific server on the inside home network As you may have guessed. The assignment usually occurs when the DHCP-configured machine boots up or regains connectivity to the network. Port forwarding is a method of counteracting this. The DHCP server then replies to the client PC with its assigned IP address. You’ll probably have to allow access for the entire remote DHCP subnet for a particular TCP/UDP port. Configuring firewall rules to accommodate access from machines that receive their IP addresses via DHCP is therefore more difficult because the remote IP address will vary from time to time. The router/firewall has no way of telling which of the many home PCs behind it should receive the relayed data because the mapping mentioned earlier doesn’t exist beforehand. port forwarding is one of the most common methods used to host Web sites at home with DHCP DSL. and default gateway information. The assignment of the IP address usually expires after a predetermined period of time.

qxd 2/25/05 9:55 AM Page 20 20 Introduction to Networking Chapter 2 Most home router/firewalls are configured in the factory to be DHCP servers for your home network. “Configuring the DHCP Server. As you can imagine. once it has a fixed IP address. 4. Most home computers get the IP address of their DNS server via DHCP from their router/firewall. Each server in the chain will store the most frequent DNS name to IP address lookups in a memory cache that helps speed up the response. “Dynamic DNS. The router/firewall then redirects the DNS queries from your computer to the DNS name server of your ISP. You can also look at Chapter 8.” explains how to configure DNS for a Web site housed on a DHCP DSL circuit where the IP address constantly changes.” explains how you can make your Linux box into a caching or regular DNS server for your network or Web site if your ISP provides you with fixed IP addresses. Home router/firewall providing DHCP services often provides its own IP address as the DNS name server address for home computers. Here is a step-by-step description of what happens with a DNS lookup. which was created for this type of scenario. which then redirects the query to the linuxhomenetworking. . How DNS Links Your IP Address to Your Web Domain The domain name system (DNS) is a worldwide server network used to help translate easy-to-remember domain names like www. You can also make your Linux box into a DHCP server.com domain name server then responds with the IP address for www. The linuxhomenetworking. 2.com into an IP address that can be used behind the scenes by your computer.com domain’s name server.linuxhomenetworking. You can check Chapter 3 for how to configure your Linux box to get its IP address via DHCP. 5. Chapter 19. It explains the auxiliary DNS standard called dynamic DNS (DDNS). The most commonly used form of DSL will also assign the outside interface of your router/firewall with a single DHCP-provided public IP address. 6. this process can cause a noticeable delay when you are browsing the Web.” to find out how to make your Linux box provide the DHCP addresses for the other machines on your network. Your ISP’s DNS server then probably redirects your query to one of the 13 root name servers. “Configuring DNS. 1.com.linuxhomenetworking.com DNS name servers. The root server then redirects your query to one of the Internet’s . 3.03Harrison_ch02. Chapter 18.

This means you can then use IP addresses from #1 to #254 on your private network. A simple analogy would be a phone number. Each 255 means this octet is for the area code (network portion).255. such as regular Web surfing and chat. So if your server has an IP address of 192. IPv6 also inherently supports the IPSec protocol suite used in many forms of secured networks. over less time-sensitive ones.0 would be the network address and 192. this book uses the latter term for the sake of consistency. billion.0. such as virtual private networks (VPNs).1. HOW SUBNET MASKS GROUP IP ADDRESSES INTO NETWORKS Subnet masks are used to tell which part of the IP address represents the network on which the computer is connected (network portion) and the computer’s unique identifier on that network (host portion). such as (808) 225-2468. Expect it to become an increasingly bigger part of your network planning in years to come. . Version 6.0. which uses a much larger 128-bit address.1.1 and the server or host would be device #25 on that network.168. which it is hoped should last for most of the 21st century.03Harrison_ch02. and the 225-2468 represents the telephone within that area code. Most home networks use a subnet mask of 255.255.168. IPv6 packets are also labeled to provide quality-of-service information that can be used in prioritizing real-time applications. offers 80 billion.255 would be used for broadcasts. the allocation of version 4 addresses will soon be exhausted between all the world’s ISPs. the network portion would be 192. which I have described here. The (808) represents the area code.255. You should also be aware that there is now a version 6 (IPv6) that has recently been developed as a replacement. the first IP address in a network is reserved as the network’s base address and the last one is reserved for broadcast traffic that is intended to be received by all devices on the network.qxd 2/25/05 9:55 AM Page 21 How Subnet Masks Group IP Addresses into Networks 21 IP Version 6 (IPv6) Most Internet-capable networking devices use version 4 of the Internet Protocol (IPv4). In all cases.168. billion times more IP addresses. With only 32 bits. Most current operating systems support IPv6 even though it isn’t currently being used extensively within corporate or home environments. In our example. Subnet masks enable you to specify how long you want the area code to be (network portion) at the expense of the number of telephones that are in the area code (host portion).25 and a subnet mask of 255. The term netmasks is often used interchangeably with the term subnet masks.168.1. such as video and voice.255. 192.

248.255. Table 2.255.255.255.255. Subtract the last octet of the subnet mask from 256 to give the number of IP addresses in the subnet (256 – 248 = 8). 224.240 255.255. Divide the last octet of the IP address by the result of step 1.255.128 255. A list of the most commonly used masks in the office or home environment is presented in Table 2. Calculating the Range of Addresses on Your Network If someone gives you an IP address of 97.255.192 255. 192.255. This gives you the theoretical number of subnets of the same size that are below this IP address. but in cases where less than 255 IP addresses are required a few rules apply. the boundaries of your network? The following section outlines the steps to do this using both a manual and programmed methodology.2 The Dotted Decimal and Slash Subnet Mask Notations Slash Format /24 /25 /26 /27 /28 /29 /30 Available Addresses 256 128 64 32 16 8 4 Dotted Decimal Format 255. so I’ll not discuss larger networks because that can become complicated.255.255. 28 / 8 = 3). 240.0 may be called a /24 mask instead.255. Manual Calculation Take out your pencil and paper—manual calculation can be tricky. . This is not a pure networking text.255. but rather by the actual number of bits in the mask.255. So for example a mask of 255. You can calculate the number of IP addresses for each of these by subtracting the value from 256.qxd 2/25/05 9:55 AM Page 22 22 Introduction to Networking Chapter 2 Calculating the Number of Addresses Assigned to a Subnet Most office and home networks use networks with 255 IP addresses or less in which the subnet mask starts with the numbers 255. 2.255. 248 and 252. you have 64 IP addresses in your subnet (256 – 192). 128.255.248 255. if you have a subnet mask of 255.255.03Harrison_ch02.28 and a subnet mask of 255.252 So for example.158.192. how do you determine the network address and the broadcast address—in other words. There are only seven possible values for the last octet of a subnet mask. These are 0.255. don’t bother with the remainder (for example.255.2.255.0 255.255.224 255. Here we go! 1.253.255. In many cases the subnet mask isn’t referred to by the dotted decimal notation.

255.0. The broadcast address is 97.0.qxd 2/25/05 9:55 AM Page 23 How Subnet Masks Group IP Addresses into Networks 23 3.0. the last octet of your network base address must be divisible by the “256 minus the last octet of your subnet mask” and leave no remainder.240: 1. 64 + 16 –1 = 79. It will accept subnet masks in dotted decimal format or /value format: [root@bigboy tmp]# . Calculation Using a Script There is a BASH script in Appendix II that will do this for you. 2. The network address is therefore 97. If you are sub-netting a large chunk of IP addresses it’s always a good idea to lay it out on a spreadsheet to make sure there are no overlapping subnets. Multiply this result by the result of step 1 to get the network address (8 × 3 = 24).193. Let’s do this for 10. (24 + 8 –1 = 31). 3.03Harrison_ch02.32. 4.92 Network Base Address : 216.168. Note As a rule of thumb.193.0. 32 + 32 – 1 = 63. 256 – 224 = 32 56 / 32 = 1 32 × 1 = 32. Therefore the network base address is 10.0. 256 – 240 = 16 75 / 16 = 4 16 × 4 = 64. Think of it as the third subnet with 8 addresses in it.31.168. 4.255”.95 Subnet Mask Subnet Size [root@bigboy tmp]# : 255.158.255.3. Let’s do this for 192. 3.64. The broadcast address is the result of step 3 plus the result of step 1 minus 1.253.158.151.255.75 with a mask of 255. this calculation exercise only works with subnet masks that start with “255. Here is an example of how to use it.sh 216. 2.151.92 /28 IP Address : 216.193.255.255.224: 1.0.151.79.3. 4. just provide the IP address followed by the subnet mask as arguments.255.168.193.240 : 16 IP Addresses .3. Once again. Think of it as the broadcast address being the network address plus the number of IP addresses in the subnet minus 1./subnet-calc.255.80 Broadcast Address : 216. Therefore the broadcast address is 192.56 with a mask of 255. Therefore the network base address is 192. Therefore the broadcast address is 10.151.253.24.63.

and error-control methodologies of data communication. but knowledge of the first two layers of the OSI model is important.253. also known as collisions. Ethernet used to operate primarily in a mode where every computer on a network section shared the same Ethernet cable.158. They would then send their data while comparing what they wanted to send with what they actually sent on the cable as a means of error detection.253. The ability to detect garbling due to simultaneous data transmissions.25.31 Network base address Gateway Available Available Available Available Available Broadcast THE PHYSICAL AND LINK LAYERS TCP/IP can be quite interesting. You will frequently see references to Ethernet being a CSMA/CD technology for this reason. There are very many standards that define the physical.248 that defines 8 IP addresses.158.253.255.158.255.253.158. which is available in a variety of cable types and speed capabilities. and similar schemes are now being used in wireless networks.158. is called collision detect or CD. but the data transmission and error correction strategy is the same in all.253.158.255. multiple access or CSMA. even the most basic communication would be impossible.24 97.253. or cyclic redundancy check (CRC).30 97. Transmitting data only after first sensing whether the cable. detected any differences between the two.26 97. your IP addresses will be 97.158.27 97.253. because without them. if the ISP provides you with a public network address of 97.158.29 97. they will most likely provide you with a subnet mask of 255.253.158. had the correct signaling levels is a methodology called carrier sense. Computers would wait until the line was clear before transmitting.qxd 2/25/05 9:55 AM Page 24 24 Introduction to Networking Chapter 2 Subnet Masks for the Typical Business DSL Line If you purchased a DSL service from your ISP that gives you fixed IP addresses. . the server would assume that it transmitted data simultaneously with another server on the cable.28 97. If a mathematical comparison. too. One of the most popular ones in departmental networks is Ethernet.255.248. electrical. which was strung between multiple devices. a subnet mask of 255. and a gateway of 97.253.253.25 97.158.03Harrison_ch02.24. It would then wait some random time and retransmit at some later stage when the line was clear again. For example.

11 specifications that define many wireless networking technologies are another example of commonly used layer 1 and 2 components of the OSI model. Now we’ll cover the hardware used to implement them.qxd 2/25/05 9:55 AM Page 25 Networking Equipment Terminology 25 Ethernet devices are now usually connected via a dedicated cable. the most common types of NIC used in the home and office are Ethernet and wireless Ethernet cards. DSL. Devices that want to transmit information have to wait their turn until the coast is clear. Duplex Explained Full duplex data paths have the capability of allowing the simultaneous sending and receiving of data. The original Ethernet standard has a speed of 10Mbps. and T1 circuits are all parts of these layers. the most recent versions can handle up to 40Gbps! The 802. This indicates that you are using the correct type of cable and that the duplex has been negotiated correctly between the devices at both ends. The Meaning of the NIC Link Light The link light signifies that the NIC card has successfully detected a device on the other end of the cable. Network Interface Cards Your network interface card is also frequently called a NIC. Full duplex uses separate pairs of wires for transmitting and receiving data so that incoming data flows don’t interfere with outgoing data flows. Currently. using more powerful hardware capable of simultaneously transmitting and receiving without interference. Half duplex data paths can transmit in both directions too. but in only one direction at a time. NETWORKING EQUIPMENT TERMINOLOGY Up to this point you have had only an introduction to the theory of the first two OSI layers. at which point they send the data. . The next few sections describe many physical and link layer concepts and the operation of the devices that use them to connect the computers in our offices and homes. Half duplex uses the same pairs of wires for transmitting and receiving data. Error-detection and dataretransmission mechanisms ensure that the data reaches the destination correctly and are specifically designed to remedy data corruption caused when multiple devices start transmitting at the same time. thereby making it more reliable and inherently faster than its predecessor versions.03Harrison_ch02. cable modem standards.

168. thereby ensuring that the packet is sent to the correct device.1.168. Most modern network cards can autonegotiate duplex with the device on the other end of the wire. in which both parties can speak at the same time. As the packet hops across the Internet.qxd 2/25/05 9:55 AM Page 26 26 Introduction to Networking Chapter 2 A good analogy for full duplex communications is the telephone. As each router in the path receives the packet. It is for this reason that duplex settings aren’t usually a problem for Linux servers. it plucks the IP packet out of the Ethernet frame. the server sends the IP packet to its NIC and tells the NIC to encapsulate the packet in a frame destined for the MAC address of the router. the packet is sent and the ARP table is subsequently updated with the new MAC address.1. If there is an ARP entry. When a server needs to communicate with another server it takes the following steps: 1. the server issues an ARP request asking that router 192. 2. The server first checks its routing table to see which router provides the next hop to the destination network. its source/destination IP address stays the same. If there is no ARP entry. on the other hand. Half duplex. 5. Every IP packet is sent out of your NIC wrapped inside an Ethernet frame that uses MAC addresses to direct traffic on your locally attached network. let’s say with an IP address of 192. Data transfer speeds will be low and error levels will be high if you have a device at one end of a cable set to full duplex and a device at the other end of the cable set to half duplex. The MAC Address The media access control (MAC) address can be equated to the serial number of the NIC. 3. If there is a valid router. is more like a walkie-talkie in which both parties have to wait until the other is finished before they can speak. but the MAC addresses are reassigned by each router on the way using a process called ARP. MAC addresses therefore have significance only on the locally attached network. the server checks its ARP table to see whether it has the MAC address of the router’s NIC. 4. How ARP Maps the MAC Address to Your IP Address The Address Resolution Protocol (ARP) is used to map MAC addresses to network IP addresses.1. You could very loosely view this as the server trying to find the Ethernet serial number of the next hop router on the local network. It then inspects the destination IP address in the packet and uses its routing . When a reply is received.03Harrison_ch02. leaving the MAC information behind.1 respond with its MAC address so that the delivery can be made.

7. . Nowadays most PCs have their COM and Ethernet ports configured as if they were going to be connected to a modem or other type of purely networking-oriented equipment. an ARP request is made asking the target server for its MAC address. The router then uses the “ARP-ing” process to get the MAC address of this next hop router. Data Terminal Equipment DTE devices were originally computer terminals located at remote offices or departments that were directly connected modems. If it doesn’t. If no entry is available. table to determine the IP address of the next router on the path to this destination.03Harrison_ch02.qxd 2/25/05 9:55 AM Page 27 Networking Equipment Terminology 27 6. As can be expected. and Data Terminal Equipment (DTE). the packet is sent and the ARP table is subsequently updated with the new MAC address. shows how to see your ARP table and the MAC addresses of your server’s NICs. It then reencapsulates the packet in an Ethernet frame with the new MAC address and sends the frame to the next hop router. 8. The server will not send the data to its intended destination unless it has an entry in its ARP table for the next hop. which covers Linux network topics. This relaying process continues until the packet reaches the target computer. the ARP table contains only the MAC addresses of devices on the locally connected network. 9. which is intended to act as the primary communications path. a similar process occurs. Data Communications Equipment DCE is also known as Data Circuit-Terminating Equipment and refers to such equipment as modems and other devices designed primarily to provide network access. The terminals would have no computing power and only functioned as a screen/keyboard combination for data processing. The Two Broad Types of Networking Equipment There are two main types of networking equipment: Data Communications Equipment (DCE). ARP entries are not permanent and are erased after a fixed period of time depending on the operating system used. which acts as the source or destination of the transmitted data. If the target server is on the same network as the source server. Chapter 3. The ARP table is queried. the application needing to communicate will issue a timeout or time exceeded error. When a reply is received.

until the message gets through correctly. or very little. electronics inside and therefore do not regulate traffic. which causes all traffic sent from a server to the hub to be blurted out to all other servers connected to that hub whether they are the intended recipient or not.03Harrison_ch02. you need a straight-through cable. When connecting two PCs (DTE) together via their COM ports.3 provides some good rules of thumb. The same would go for positions #2 through #8— that is. When connecting a PC’s COM port (DTE) to a modem (DCE) the straight-through cable is frequently called a modem cable. When this happens the servers try again. Hubs physically cross-connect all their ports with one another. Hubs have no. Hold the connectors side by side. A straight-through Ethernet cable is easy to identify. It is possible for multiple servers to speak at once. the crossover cable is often referred to as a null modem cable. As you can see. try using the other type of cable. If you fail to get a link light when connecting your Ethernet devices together. This terminology is generally used with Ethernet cables. Table 2. A crossover cable has them mixed up.qxd 2/25/05 9:55 AM Page 28 28 Introduction to Networking Chapter 2 Using Straight-Through/Crossover Cables to Connect DTEs and DCEs When a DCE is connected to a DTE.3 Cabling Rules of Thumb Likely Cable Type Crossover Crossover Crossover Straight through Straight through Straight through Scenario PC to PC Hub to hub Switch to switch PC to modem PC to switch PC to hub Connectivity Using Hubs A hub is a device into which you can connect all devices on a network so that they can talk together. Table 2. pointing in the same direction with the clips facing away from you. Some manufacturers configure the Ethernet ports of their networking equipment to be either of the DTE or the DCE type. and other manufacturers have designed their equipment to flip automatically between the two types until it gets a good link. the same color for corresponding wires on each end. The terminology can be different for cables used to connect serial ports together. with all of them receiving garbled messages. after a random time interval. . confusion can arise when selecting a cable. The color of the wire in position #1 on connector #1 should be the same as that of position #1 on connector #2. DCEs connected to DCEs or DTEs connected to DTEs require crossover cables.

It is possible to connect multiple switches and/or hubs in a chain formation to create a LAN with more ports. Switches regulate traffic. Unlike a hub. thereby eliminating the possibility of message garbling and providing a more efficient traffic flow. By connecting its NIC cards to multiple LANs. This is often called daisy chaining. Devices that plug into switches should be set to full duplex to take full advantage of the dedicated bandwidth coming from each switch port. Local Area Networks A local area network (LAN) is a grouping of ports on a hub. a correctly configured router is capable of relaying traffic between networks.qxd 2/25/05 9:55 AM Page 29 Networking Equipment Terminology 29 It is for these reasons that Ethernet devices that plug into hubs should be set to half duplex. Note Hubs can add a lot of delays to your network because of the message garbling collisions and retransmissions. they attempt to communicate with each other via routers that interconnect their LANs. A switch is a much more reliable and predictable alternative. Larger. therefore providing additional security. . switches and hubs usually have only servers connected to them that have been configured as being part of the same network. A router still needs to be connected to each VLAN for internetwork communication. The only exception is broadcast traffic. Routers are also capable of filtering traffic passing between the two LANs. Servers use their IP address and subnet mask and the IP address of the remote server to determine whether they are both on the same network. the switch houses ports on multiple LANs. In this case. If not. This is why network administrators group trusted servers having similar roles on the same LAN. more expensive switches can be configured to assign only certain ports to prespecified virtual LANs (VLANs) chosen by the network administrator. and ones made for the home often cost only a few dollars more.03Harrison_ch02. switch. Switches and hubs provide no access control between servers connected to the same LAN. How Routers Interconnect LANs As stated before. or tied to a wireless access point (WAP) that can communicate only with each other. which is blurted out to all the servers simultaneously. traffic sent from Server A to Server B will be received only by Server B. Using Switches as a Faster Alternative to Hubs A switch is also a device into which you can connect all devices on a home network so that they can talk together.

especially if only one network protocol is being used. In other words. If you intend to route between networks. How Simple Routing Works In the broader networking sense.” Note The term gateway specifically refers to a device that routes traffic between dissimilar network protocols (IP to Appletalk) or access methods (Ethernet to DSL). much like a traffic policeman. a “route” refers to the path data takes to traverse from its source to its destination. routers most frequently provide connectivity to the Internet using network address translation or NAT. The terms are frequently used interchangeably. Routers are designed to exchange routing information dynamically and can therefore intelligently redirect traffic to bypass failed network links. Therefore a home DSL router that provides IP Internet access to an Ethernet network is technically a gateway. shows how to add static routes to your Linux box and also how you can convert it into a simple router. we are referring to the IP address of the first hop needed to reach the desired destination network. Routers therefore direct and regulate traffic between separate networks. routers act as gateways to the wider world and it won’t be surprising to learn that routers are frequently referred to as “gateways.qxd 2/25/05 9:55 AM Page 30 30 Introduction to Networking Chapter 2 Routers can also be configured to deny communication between specific servers on different networks. Each router along the way may also be referred to as a hop. The distinction can be important in complicated networking environments where newer technologies need to talk with older ones using incompatible communications protocols. They can also filter traffic based on the TCP port section of each packet. In home networks. Routers transfer traffic where both the protocols and communications medium are the same. you must reserve an IP address for a router for each network and make sure that the router is directly connected to the LAN associated with that network. It is assumed that this first hop will know how to automatically relay the packet. which covers Linux network topics.03Harrison_ch02. Usually when we speak about a route on a Linux box. . For example. Home Linux boxes frequently don’t run a dynamic routing protocol and therefore rely on static routes issued by the system administrator at the command line or in configuration files to determine the next hop to all desired networks. it is possible to deny communication between two servers on different networks that intend to communicate on TCP port 80 and allow all other traffic between them. Chapter 3.

the Routers of Last Resort A default gateway is the router that is used when no alternative devices can be found to relay the traffic. R1 and R2.0 255.qxd 2/25/05 9:55 AM Page 31 Networking Equipment Terminology 31 Default Gateways.0. . shows how to configure the default gateway on your Linux box.0. ☞ Go to everything else via router R1.X via another NIC card. your default gateway would be the router/ firewall connected to the Internet. They are often called routers of last resort. Firewalls Help Provide a Secure Routing Environment Firewalls can be viewed as routers with more enhanced abilities to restrict traffic. These VPNs are frequently used to connect branch offices to the corporate headquarters and also to allow sales representatives to get access to sensitive pricing information when traveling from town to town. not just by port and IP address as routers do. Specifically. Firewalls can often create an encrypted data path between two private networks across the Internet providing secure communication with a greatly reduced chance of eavesdropping. Say for example you have two routers.0. For most home networks. firewalls can detect malicious attempts to subvert the TCP/IP protocol. R1 is connected to both your SOHO home network and the Internet.X. Chapter 3.0. Firewalls are used as close to the source or target of data communication as possible to try to ensure that the data hasn’t been subverted. You could put a route on your SOHO servers that states: ☞ Go to network 10.03Harrison_ch02.0 via router R2. which covers Linux network topics. A short list of capabilities includes the following: ☞ Throttling traffic to a server when too many unfulfilled connections are made to it ☞ Restricting traffic being sent to obviously bogus IP addresses ☞ Providing network address translation or NAT Routers are designed to make packets flow as quickly as possible with the minimum amount of inspection. R1 therefore would be considered your default gateway. R2 is connected to SOHO home network and is capable of relaying data to other corporate networks with addresses starting with 10.X.

qxd 2/25/05 9:55 AM Page 32 32 Introduction to Networking Chapter 2 ADDITIONAL INTRODUCTORY TOPICS The last few topics of this chapter may not appear to be directly related to networking. You can also use FTP from the command line. which gives a list of all the man pages that contain the word ssh: . The remote FTP server prompts you for a username. such as WSFTP and CuteFTP. there are two types of FTP. The File Transfer Protocol FTP is one of the most popular applications used to copy files between computers via a network connection. you use the command man ssh: [root@bigboy tmp]# man ssh If you want to search all the man pages for a keyword. From the systems administrator’s perspective. “Troubleshooting Linux with Syslog. The second method. From the command line you issue the man command followed by the Linux command or file about which you want information. From the remote user’s perspective. and data are sent across the network unencrypted. use the man command with the -k switch—for example. but they cover Linux help methods that you’ll use extensively and the File Transfer Protocol (FTP) package. anonymous FTP. as shown in Chapter 5. These are active and passive FTP.03Harrison_ch02. there are two more categories. Linux Help Linux help files are accessed using the man or manual pages. The first is regular FTP.” Remember that FTP isn’t very secure because usernames. The remote FTP server prompts you for a specific username and password to gain access to the data. Knowledge of FTP is especially important and is a primary method of downloading software for Linux systems. is used primarily to allow any remote user to download files to their systems. at which point the user types anonymous or ftp with the password being any valid e-mail address. which enables you to download all the software you need to get your Linux server operational as quickly as possible. There are a number of commercially available GUI-based clients you can load on your PC to do this. “Linux FTP Server Setup. which is used primarily to allow specific users to download files to their systems.” on RPM software installation. “Secure Remote Logins and File Copying”) that is normally installed by default with Fedora. If you want to get information on the ssh command. passwords. man -k ssh. which are covered in more detail in Chapter 15. More secure forms—such as SFTP (Secure FTP) and SCP (Secure Copy)—are available as a part of the Secure Shell package (covered in Chapter 17.

. use its man page to get more details on what it does and the syntax it needs. .. .. it’s time to read Chapter 3 to learn how to configure Linux networking. Feel free to return to this chapter whenever you need to refresh your memory on these foundation concepts. ssh (1) ssh [slogin] (1) ssh-agent (1) ssh-keyscan (1) ssh_config (5) sshd (8) sshd_config (5) . so I’ll be using a wide variety of commands in this book without detailed explanations to help keep the flow brisk. [root@bigboy tmp]# ssh - OpenSSH SSH client (remote login OpenSSH SSH client (remote login authentication agent gather ssh public keys OpenSSH SSH client configuration OpenSSH SSH daemon OpenSSH SSH daemon configuration program) program) files file This book is targeted at proficient Linux beginners and above. If you need more help on a command.qxd 2/25/05 9:55 AM Page 33 Conclusion 33 [root@bigboy tmp]# man -k .... . CONCLUSION Congratulations! Now that you have an understanding of basic networking...03Harrison_ch02. but with a little practice the man pages can become your friend. Linux help can sometimes be cryptic.

03Harrison_ch02.qxd 2/25/05 9:55 AM Page 34 .

Some of these activities are automatically covered during a Linux installation. or use an alternative means of connecting to the Internet. add a new network interface card. but you will often find yourself having to know how to modify these initial settings whenever you need to move your server to another network. 35 . In Chapter 2.qxd 2/25/05 9:56 AM Page 35 C H A P T E R 3 Linux Networking IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ How to Configure Your NIC’s IP Address How to Activate/Shut Down Your NIC How to View Your Current Routing Table How to Change Your Default Gateway How to Configure Two Gateways How to Delete a Route Changing NIC Speed and Duplex How to Convert Your Linux Server Into a Router Configuring Your /etc/hosts File Conclusion Now that you have a firm grasp of many of the most commonly used networking concepts.” we started with an explanation of TCP/IP.04Harrison_ch03. it is time to apply them to the configuration of your server. so we’ll start this Linux networking chapter with a discussion on how to configure the IP address of your server. “Introduction to Networking.

0.255.1.255.0 b) TX bytes:0 (0.168.qxd 2/25/05 9:56 AM Page 36 36 Linux Networking Chapter 3 HOW TO CONFIGURE YOUR NIC’S IP ADDRESS You need to know all the steps to configure IP addresses on a NIC card. This section shows you how to do the most common server IP activities with the least amount of headaches.04Harrison_ch03. you might just want to play around with the server to test your skills.255. Web site shopping cart applications frequently need an additional IP address dedicated to them.1 Mask:255.168.255 Mask:255. You can determine the IP address of this device with the ifconfig command: [root@bigboy tmp]# ifconfig -a eth0 Link encap:Ethernet HWaddr 00:08:C7:10:74:A8 BROADCAST MULTICAST MTU:1500 Metric:1 RX packets:0 errors:0 dropped:0 overruns:0 frame:0 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:0 (0.0 b) Interrupt:11 Base address:0x1820 lo Link encap:Local Loopback inet addr:127.1.7 Kb) wlan0 Link encap:Ethernet HWaddr 00:06:25:09:6A:B5 inet addr:192.168.255 Mask:255. this device is called eth0.1.168.255.1. Determining Your IP Address Most modern PCs come with an Ethernet port.4 Mb) TX bytes:43209032 (41.100 Bcast:192.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:47379 errors:0 dropped:0 overruns:0 frame:0 TX packets:107900 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:4676853 (4.0.0.0.99 Bcast:192. When Linux is installed.7 Kb) TX bytes:82644 (80.2 Mb) Interrupt:11 Memory:c887a000-c887b000 wlan0:0 Link encap:Ethernet HWaddr 00:06:25:09:6A:B5 inet addr:192. Last but not least.0 UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:787 errors:0 dropped:0 overruns:0 frame:0 TX packets:787 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:82644 (80. You also might need to add a secondary NIC interface to your server to handle data backups.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 Interrupt:11 Memory:c887a000-c887b000 [root@bigboy tmp]# .

wlan0 has an IP address of 192. and so on.168.04Harrison_ch03.0 up The up at the end of the command activates the interface. You can place your IP address information in these files. which is run at the end of every reboot. See Figure 3.0. or PCI bus ID.1 netmask 255. eth0 has no IP address because this box is using wireless interface wlan0 as its main NIC Interface. Devices eth0 and eth1 use interrupts 10 and 5.100 and a subnet mask of 255. you could give this ifconfig command: eth0 interface an IP address using the [root@bigboy tmp]# ifconfig eth0 10.255. . One assumes the interface has a fixed IP address. In the example below we can see that there are no conflicts with each IRQ from 0 to 15 having only a single entry. which are then used to autoconfigure your NICs when Linux boots. On very rare occasions you may find that your NIC card doesn’t work because it shares both an interrupt and memory access address with some other device.255.1 for two samples of interface eth0.local file. respectively: [root@bigboy tmp]# cat /proc/interrupts CPU0 0: 2707402473 XT-PIC timer 1: 67 XT-PIC i8042 2: 0 XT-PIC cascade 5: 411342 XT-PIC eth1 8: 1 XT-PIC rtc 10: 1898752 XT-PIC eth0 11: 0 XT-PIC uhci_hcd 12: 58 XT-PIC i8042 14: 5075806 XT-PIC ide0 15: 506 XT-PIC ide1 NMI: 0 ERR: 43 [root@bigboy tmp]# If there are conflicts.qxd 2/25/05 9:56 AM Page 37 How to Configure Your NIC’s IP Address 37 In this example.1. you might need to refer to the manual for the offending device to try to determine ways to either use another interrupt or memory I/O location. Fedora Linux also makes life a little easier with interface configuration files located in the /etc/sysconfig/network-scripts directory. You can look at the contents of the /proc/interrupts file to get a listing of all the interrupt IRQs used by your system. eth1 uses ifcfg-eth1. and the other assumes it requires an IP address assignment using DHCP. used by each card. Interface eth0 has a file called ifcfg-eth0.255.0. Changing Your IP Address If you wanted. To make this permanent each time you boot up you have to add this command in your /etc/rc.255.0 You can see that this command gives good information on the interrupts.

eth0 will be activated on booting.1 File formats for network-scripts.255.255. because the parameter has the value yes and not no. The default Red Hat/Fedora installation will include the broadcast and network options in the network-scripts file. These are optional.qxd 2/25/05 2:33 PM Page 38 38 Linux Networking Chapter 3 Fixed IP Address [root@bigboy tmp]# cd /etc/sysconfig/network-scripts [root@bigboy network-scripts]# less ifcfg-eth0 DEVICE=eth0 IPADDR=192. You can read more about netmasks and DHCP in Chapter 2.1.168. The ifdown and ifup commands can be used to do this: ONBOOT [root@bigboy network-scripts]# ifdown eth0 [root@bigboy network-scripts]# ifup eth0 Your server will have to have a default gateway for it to be able to communicate with the Internet.168.168.04Harrison_ch03. which acts as an introduction to networking.1. As you can see.255 NETWORK=192.1.0 [root@bigboy network-scripts]# Getting the IP Address Using DHCP [root@bigboy tmp]# cd /etc/sysconfig/network-scripts [root@bigboy network-scripts]# less ifcfg-eth0 DEVICE=eth0 BOOTPROTO=dhcp ONBOOT=yes [root@bigboy network-scripts]# Figure 3. This will be covered later in the chapter. After you change the values in the configuration files for the NIC you have to deactivate and activate it for the modifications to take effect. .100 NETMASK=255.0 BOOTPROTO=static ONBOOT=yes # # The following settings are optional # BROADCAST=192.

a virtual subinterface also known as an IP alias.04Harrison_ch03.168.99 \ netmask 255. Here is a sample configuration: DEVICE=wlan0:0 ONBOOT=yes BOOTPROTO=static IPADDR=192.168. “Determining Your IP Address. but also the desired DNS servers. You should also create a /etc/sysconfig/network-scripts/ifcfg-wlan0:0 file so that the aliases will all be managed automatically with the ifup and ifdown commands. 2.255. The process for creating an IP alias is very similar to the steps outlined for the real interface in the previous section.0 The commands to activate and deactivate the alias interface would therefore be: [root@bigboy tmp]# ifup wlan0:0 [root@bigboy tmp]# ifdown wlan0:0 Note Shutting down the main interface also shuts down all its aliases.0 up 4. In this we want to create interface wlan0:0. Aliases can be shut down independently of other interfaces.1.1.255. Create the virtual interface with the ifconfig command: [root@bigboy tmp]# ifconfig wlan0:0 192.qxd 2/25/05 9:56 AM Page 39 How to Configure Your NIC’s IP Address 39 How DHCP Affects the DNS Server You Use Your DHCP server not only supplies the IP address your Linux box should use. Make sure your /etc/resolv. First ensure the parent real interface exists. Multiple IP Addresses on a Single NIC In the previous section.255. “Changing Your IP Address:” 1.” you may have noticed that there were two wireless interfaces: wlan0 and wlan0:0. Aliases have the name format parent-interface-name:X.255. After completing these four simple steps you should be able to ping the new IP alias from other servers on your network. where X is the subinterface number of your choice. Interface wlan0:0 is actually a child of interface wlan0. . IP aliasing is one of the most common ways of creating multiple IP addresses associated with a single NIC. Verify that no other IP aliases with the same name exist with the name you plan to use.conf file has the servers configuration lines commented out to prevent any conflicts. 3.99 NETMASK=255.

ifcfg-eth0 2. “Installing RPM Software. After installing the RPM. subnet mask.04Harrison_ch03. When searching for the file. the process is different. If you are using a DSL connection with a DHCP or dynamic IP address assignment.qxd 2/25/05 9:56 AM Page 40 40 Linux Networking Chapter 3 IP Address Assignment for a Direct DSL Connection If you are using a DSL connection with fixed or static IP addresses. Your ISP will provide you with a PPP authentication over Ethernet (PPPoE) username and password that will allow your computer to log in transparently to the Internet each time it boots up.i386. The DHCP IP address assigned to your computer and/or Internet router will often not change for many days and you may be fooled into thinking it is static. DEVICE=eth0 ONBOOT=no 3. your ISP will provide you with a DHCP-based connection. broadcast address. Downloading and installing RPMs isn’t hard. Make a backup copy of your ifcfg-eth0 file: [root@bigboy tmp]# [root@bigboy tmp]# cd /etc/sysconfig/network-scripts/ [root@bigboy network-scripts]# ls ifcfg-eth0 ifcfg-eth0 [root@bigboy network-scripts]# cp ifcfg-eth0 DISABLED. Shut down your eth0 interface: [root@bigboy network-scripts]# ifdown eth0 [root@bigboy network-scripts]# .rpm. configure it with the IP address. Edit your ifcfg-eth0 file to have no IP information and also to be deactivated on boot time. If you need a refresher. Fedora Linux installs the rp-pppoe RPM software package required to support this. remember that the PPPoE RPM’s filename usually starts with the word rp-pppoe followed by a version number like this: rp-pppoe3. and gateway information provided by your ISP and you should have connectivity when you restart your interface.” covers how to do this in detail.5-8. Chapter 6. the configuration steps are the same as those outlined earlier. The PPPOE configuration will create a softwarebased virtual interface named ppp0 that will use the physical Internet interface eth0 for connectivity. you need to go through a number of steps to complete the connection. You plug your Ethernet interface into the DSL modem. Note Unless you specifically request static IP addresses. Remember that you may also need to configure your DNS server correctly. Here’s what you need to do: 1.

this is likely to be something like /dev/hme0. First. enter the word server. I will run some checks on your system to make sure the PPPoE client is installed properly. enter ‘no’ (two letters.” for more information on BIND and DNS.. If you want your ISP to provide the IP address of its DNS server automatically. If you’re running BIND on your server in a caching DNS mode then you may want to leave this option blank.04Harrison_ch03. We’ll use defaults wherever possible: Welcome to the ADSL client setup. Enter the demand value (default no): It will then prompt you for your DNS server information. the interface to be used (eth0). lower-case. Run the adsl-setup configuration script: [root@bigboy network-scripts]# adsl-setup It will prompt you for your ISP username. For Linux. If your ISP claims that ‘the server will provide dynamic DNS addresses’. enter the idle time in seconds after which the link should be dropped. “Configuring DNS. Please refer to Chapter 18.. If you just press enter.) NOTE: Demand-activated links do not interact well with dynamic IP addresses.conf file. DNS Please enter the IP address of your ISP’s primary DNS server. (default eth0): Do you want the link to come up on demand. and whether you want the connection to stay up indefinitely. where ‘X’ is a number.qxd 2/25/05 9:56 AM Page 41 How to Configure Your NIC’s IP Address 41 4. LOGIN NAME Enter your Login Name (default root): bigboy-login@isp INTERFACE Enter the Ethernet interface connected to the ADSL modem For Solaris. If you want the link to stay up permanently. Enter the DNS information here: The script will then prompt you for your ISP password: PASSWORD Please enter your Password: Please re-enter your Password: . This step edits your /etc/resolv. You may have some problems with demand-activated links. enter ‘server’ (all lower-case) here. I will assume you know what you are doing and not modify your DNS setup. it will be ethX. or stay up continuously? If you want it to come up on demand.

The firewall choices are: 0 .STANDALONE: Appropriate for a basic stand-alone web-surfing workstation 2 . Note that these rules are very basic. DNS server. “Linux Firewalls Using iptables. This may be required if non-root members of your family or home office need to get access to the Internet: USERCTRL Please enter ‘yes’ (three letters.qxd 2/25/05 9:56 AM Page 42 42 Linux Networking Chapter 3 Then it will ask whether you want regular users (not superuser root) to be able to activate/deactivate the new ppp0 interface. however.MASQUERADE: Appropriate for a machine acting as an Internet gateway for a LAN Choose a type of firewall (0-2): 0 You’ll then be asked whether you want the connection to be activated upon booting. They are very basic and don’t cover rules to make your Linux box a Web server. FIREWALLING Please choose the firewall rules to use.) if you want to allow normal user to start or stop DSL connection (default yes): The rp-pppoe package has two sample iptables firewall scripts located in the /etc/ppp directory named firewall-standalone and firewall-masq. e-mail. lower-case. or mail server. If you are running any servers on your machine. You are strongly encouraged to use a more sophisticated firewall setup. You are STRONGLY recommended to use some kind of firewall rules. etc. you must choose ‘NONE’ and set up firewalling yourself. the firewall rules will deny access to all standard servers like Web. 1 . You are responsible for ensuring the security of your machine. Most people would say yes: Start this connection at boot time Do you want to start this connection at boot time? Please enter no or yes (default no):yes Just before exiting.” or the more comprehensive one found in the Appendix. If you are using SSH. ftp. these will provide basic security. Otherwise. the rules will block outgoing SSH connections which allocate a privileged source port.NONE: This script will not set any firewall rules. you’ll get a summary of the parameters you entered and the relevant configuration files will be updated to reflect your choices when you accept them: . I’d recommend selecting none and using a variant of the basic script samples in Chapter 14.04Harrison_ch03.

it should be all set up! Type ‘/sbin/ifup ppp0’ to bring up your xDSL link and ‘/sbin/ifdown ppp0’to bring it down. PPPOE_TIMEOUT=20 LCP_FAILURE=3 LCP_INTERVAL=80 . and therefore listing the ifcfg-ppp0 filename won’t be necessary in most home environments. Note This example recommends using the adsl-status command with the name of the PPPoE interface configuration file. After you have completed installing rp-pppoe you should be able to access the Internet over your DHCP DSL connection as expected. The first is the ifcfg-ppp0 file with interface’s link layer connection parameters: [root@bigboy network-scripts]# more ifcfg-ppp0 USERCTL=yes BOOTPROTO=dialup NAME=DSLppp0 DEVICE=ppp0 TYPE=xDSL ONBOOT=yes PIDFILE=/var/run/pppoe-adsl. Type ‘/sbin/adsl-status /etc/sysconfig/network-scripts/ifcfg-ppp0’ to see the link status. This command defaults to show information for interface ppp0.qxd 2/25/05 9:56 AM Page 43 How to Configure Your NIC’s IP Address 43 ** Summary of what you entered ** Ethernet Interface: eth0 User name: bigboy-login@isp Activate-on-demand: No DNS: Do not adjust Firewalling: NONE User Control: yes Accept these settings and adjust configuration files (y/n)? y Adjusting /etc/sysconfig/network-scripts/ifcfg-ppp0 Adjusting /etc/ppp/chap-secrets and /etc/ppp/pap-secrets (But first backing it up to /etc/ppp/chap-secrets.04Harrison_ch03.bak) (But first backing it up to /etc/ppp/pap-secrets.pid FIREWALL=NONE PING=. Some Important Files Created by adsl-setup The adsl-setup script creates three files that will be of interest to you.bak) At the very end it will tell you the commands to use to activate/deactivate your new ppp0 interface and to get a status of the interface’s condition: Congratulations.

the interface appears to work correctly: [root@bigboy tmp]# ifup ppp0 [root@bigboy tmp]# adsl-status adsl-status: Link is up and running on interface ppp0 ppp0: flags=8051<UP.RUNNING. In this case the package has been installed but the interface hasn’t been activated: [root@bigboy tmp]# adsl-status Note: You have enabled demand-connection.04Harrison_ch03.com). adsl-status may be inaccurate. but ppp0 is down [root@bigboy tmp]# After activation.. There are some good tips there on how to avoid problems with VPN clients. [root@bigboy tmp]# For further troubleshooting information you can visit the Web site of at Roaring Penguin (www... . .roaringpenguin. rp-ppoe ..MULTICAST> mtu 1462 inet . adsl-status: Link is attached to ppp0.qxd 2/25/05 9:56 AM Page 44 44 Linux Networking Chapter 3 CLAMPMSS=1412 CONNECT_POLL=6 CONNECT_TIMEOUT=60 DEFROUTE=yes SYNCHRONOUS=no ETH=eth0 PROVIDER=DSLppp0 USER= bigboy-login@isp PEERDNS=no [root@bigboy network-scripts]# secrets The others are the duplicate /etc/ppp/pap-secrets and /etc/ppp/chapfiles with the username and password needed to login to your ISP: [root@bigboy network-scripts]# more /etc/ppp/pap-secrets # Secrets for authentication using PAP # client server secret IP addresses “bigboy-login@isp” * “password” [root@bigboy network-scripts]# Simple Troubleshooting You can run the adsl-status command to determine the condition of your connection..POINTOPOINT..

so a gateway address of 0.0 0.0 [root@bigboy tmp]# .168.0. Here is an example for interface eth0: [root@bigboy tmp]# ifdown eth0 [root@bigboy tmp]# ifup eth0 HOW TO VIEW YOUR CURRENT ROUTING TABLE The netstat -nr command will provide the contents of the routing table.0. You must have an ifcfg file in the /etc/sysconfig/network-scripts directory for these commands to work.0.0. No gateway is needed to reach your own directly connected interface.168.0.255.0.0.qxd 2/25/05 9:56 AM Page 45 How to View Your Current Routing Table 45 IP Address Assignment for a Cable Modem Connection Cable modems use DHCP to get their IP addresses.255.0 192.255 0.0.0.0.0 0.0 127.0.1.04Harrison_ch03.255.1 0.255.0.0.0 Genmask 255. the default and one to 255.0. HOW TO ACTIVATE/SHUT DOWN YOUR NIC The ifup and ifdown commands can be used respectively to activate and deactivate a NIC interface.0 0.0 are usually directly connected to the interface.0.0 Flags MSS Window irtt Iface UH 40 0 0 wlan0 U U UG 40 40 40 0 0 0 0 0 0 wlan0 lo wlan0 192.255.0 255. Your default gateway is the route with a destination address of 0.0.0.0. Server bigboy is a DHCP server in this case: [root@bigboy tmp]# netstat -nr Kernel IP routing table Destination Gateway 255.0. so you can configure your server’s Ethernet interface accordingly.255 255.1. Networks with a gateway of 0.0 seems appropriate.0.255 which is usually added on DHCP servers. ☞ In this example there are two gateways.255.255.255.0.

11.0.16.128 U UG UG U UG 172.16.192 0.255.0.0 172. This example uses a newly installed wireless interface called wlan0. most PCs would be using the standard Ethernet interface eth0: [root@bigboy tmp]# route add default gw 192. DHCP servers automatically assign a default gateway to DHCP-configured NICs.160.64 172.0 172.16.0.255.255.168.16.193 255.32 172.0 0.128 0.16.1 wlan0 In this case.0 Gateway Genmask Flags MSS Window irtt 40 40 40 40 40 40 40 40 40 40 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 eth1 eth1 eth1 eth0 eth1 eth0 eth0 eth0 lo eth1 172.0.1.135 255.0.16.16.193 255.0. This can be done with a simple command. This file is used to configure your default gateway each time Linux boots: NETWORKING=yes HOSTNAME=bigboy GATEWAY=192. static IP addresses will need to have a manually configured default gateway.255.0 0.0.16.16.224 UG 172.255.192 U 255.193 0.0.16.255.0.1.0 255.0 172.67.67.0. but NICs with configured.224 UG 172.0.67.255.96 172. you need to update your /etc/sysconfig/network file to reflect the change.qxd 2/25/05 9:56 AM Page 46 46 Linux Networking Chapter 3 ☞ In this example.16.0 127.255.131 255.1 .255.240.69.255.16.69.69.0 172.0 172. make sure that the router/firewall with IP address 192.69.0.224 UG 172.1.16.168.68.67.0. there are multiple gateways handling traffic destined for different networks on different interfaces: [root@bigboy tmp]# netstat -nr Kernel IP routing table Destination Iface 172.67.1 is connected to the same network as interface wlan0! When this is done.0.04Harrison_ch03.255.168.224 UG 255.0.135 255.0 172.68.193 255.255.16.255.0.0 [root@bigboy tmp]# HOW TO CHANGE YOUR DEFAULT GATEWAY Your server needs to have a single default gateway.0.69.

☞ You also have another router providing access to your corporate network using addresses in the range 10.0.1 gw 192. You may be most likely using interface eth0.254 wlan0 .255.255.0. There are a number of ways to add this new route.0.0. It is possible to define default gateways in the NIC configuration file in the /etc/sysconfig/network-scripts directory.255.168. Here’s a typical scenario: ☞ You have one router providing access to the Internet that you’d like to have as your default gateway (see the default gateway example earlier).0. please adjust your steps accordingly.0 \ gw 192. If one of the default gateways has no route to the intended destination.255.0.0.1.254 wlan0 If you wanted to add a route to an individual server. This will cause packets to be sent to alternating gateways and could cause connectivity problems.d/rc.255.168. and the subnet mask and gateway values also have to be preceded by the netmask and gw switches.255.254 The Linux box used in this example uses interface wlan0 for its Internet connectivity.1: [root@bigboy tmp]# route add -host 10. In our example the reference to the 10.0 netmask 255. (The route command automatically knows the mask should be 255. every other packet will become lost. HOW TO CONFIGURE TWO GATEWAYS Some networks may have multiple router/firewalls providing connectivity.qxd 2/25/05 9:56 AM Page 47 How to Configure Two Gateways 47 Some people don’t bother with this step and just place the route add command in the script file /etc/rc. respectively: [root@bigboy tmp]# route add -net 10.1.0 to 10.0.) Here is an example for a route to host 10. Let’s assume that this router has an IP address of 192.1. Adding Routes from the Command Line The route add command can be used to add new routes to your server.0.168. Firewalls that are designed to block packets with irregular sequence numbers and unexpected origins could also obstruct your data flow. the -host switch would be used with no netmask value.0.0 network has to be preceded with a -net switch.local which is run at the end of every reboot. but you run the risk of inadvertently assigning more than one default gateway when you have more than one NIC.04Harrison_ch03.0. It has the advantage of being universal to all versions of Linux and is well documented in the man pages.0.

There will also be no reported errors on the screen or any of the log files in the /var/log/ directory.255.1.0.0.1.0.0.0.04Harrison_ch03.0 255.1 [root@bigboy tmp]# -nr Genmask 255.168./ifup-routes wlan0 [root@bigboy tmp]# netstat Kernel IP routing table Destination Gateway 192.1 [root@bigboy tmp]# -nr Genmask 255.0.0 255.254 0. The format of the file is quite intuitive with the target network coming in the first column followed by the word via and then the gateway’s IP address.0 (a mask with the first 8 bits set to 1) via the 192.0.0 network and execute the ifup-routes command.0.0 192.1.0 169.0 0.0 255.0.0 0.254.0 192.255.1.0.255.0 Flags U U UG UG MSS 0 0 0 0 Window 0 0 0 0 irtt 0 0 0 0 Iface wlan0 wlan0 wlan0 wlan0 .0 0. so the filename for interface wlan0 would be route-wlan0.254 Note The /etc/sysconfig/network-scripts/route-* filename is very important.0 0.168.0 169. You can test the new file by running the /etc/sysconfig/networkcommand with the interface name as the sole argument.168.168.0.0/8 via 192.0.168.qxd 2/25/05 9:56 AM Page 48 48 Linux Networking Chapter 3 A universal way of making this change persistent after a reboot would be to place this route add command in the file /etc/rc. In the next example we check the routing table to see no routes to the 10.0.1.255. Adding the wrong interface extension at the end will result in the routes not being added after the next reboot.d/rc.1.0.0.0.0.254 gateway.0.168.0.0.0.0.0 0. In our routing example.0.local.0. which is always run at the end of the booting process.0.0 192.0 0.255.1. The filename format is route-interface-name.0.168.255.0.0 Flags U U UG MSS 0 0 0 Window 0 0 0 irtt 0 0 0 Iface wlan0 wlan0 wlan0 scripts/ifup-routes [root@bigboy tmp]# .0.0.0 with a subnet mask of 255. which then adds the route: [root@bigboy tmp]# netstat Kernel IP routing table Destination Gateway 192.0.0 10.0 0.0. we would have to configure file /etc/sysconfig/network-scripts/route-wlan0 to look like this: # # File /etc/sysconfig/network-scripts/route-wlan0 # 10. permanent static routes are added on a per interface basis in files located in the /etc/sysconfig/network-scripts directory. Adding Routes with /etc/sysconfig/network-scripts/ Files In Fedora Linux.0. to set up a route to network 10.254.

0.0. so you need to be aware of it. Don’t limit your troubleshooting of these types of errors to just failed negotiation. but only the fully compatible ones will provide you with errorfree. the problem could also be due to a bad NIC card.0 \ gw 192. but this is usually accompanied by many collision-type errors being seen on the NIC. when using the ifconfig -a command.1.168. Typically. and only marginal performance. Using mii-tool One of the original Linux tools for setting the speed and duplex of your NIC card was the mii-tool command.qxd 2/25/05 9:56 AM Page 49 Changing NIC Speed and Duplex 49 HOW TO DELETE A ROUTE Here’s how to delete the routes added in the previous section: [root@bigboy tmp]# route del -net 10.0. NICs with failed negotiation will work.1. Configuring a switch port to autonegotiate the speed and duplex often isn’t sufficient because there are frequently differences in the implementation of the protocol standard. Linux defaults to negotiating the speed and duplex of its NIC automatically with that of the switch to which it is attached. Issuing the command without any arguments gives a brief status report. or cabling. Most Linux vendors have a list of compatible hardware on their Web sites. but you have to refer to the manufacturer’s guides to set the speed and duplex to anything but auto-negotiate. the server will not reinsert the route. It is destined to be deprecated and replaced by the newer ethtool command. If you can’t find any of the desired models in your local computer store. as seen in the next example.254 wlan0 The file /etc/sysconfig/network-scripts/route-wlan0 will also have to be updated so that when you reboot. . consistent throughput.0/8 via 192.0. a model in the same family or series should be sufficient. read this carefully before you start hooking up your machine to the network.0 netmask 255.04Harrison_ch03. Delete the line that reads 10. switch port.168. but many older NICs support only mii-tool. NICs that are not compatible with mii-tool often will still work. Most cards will work.0.254 CHANGING NIC SPEED AND DUPLEX There is no better Linux investment than the purchase of a fully Linux-compatible NIC card. with unsupported NICs providing an Operation not supported message.0.

negotiation was OK. “The Linux Boot Process.04Harrison_ch03. Here we see interface eth0 not doing autonegotiation and set to a speed of 100Mbps. 100baseTx-HD. In this case. half duplex. The command provides the status of the interface you provide as its argument. Creating your own startup scripts is covered in Chapter 7. model 33 basic mode: autonegotiation enabled basic status: autonegotiation complete. capabilities: 100baseTx-FD 100baseTx-HD advertising: 100baseTx-FD 100baseTx-HD link partner: 100baseTx-FD 100baseTx-HD control [root@bigboy tmp]# rev 2 link ok 10baseT-FD 10baseT-HD 10baseT-FD 10baseT-HD 10baseT-FD 10baseT-HD flow- Setting Your NIC’s Speed Parameters with mii-tool You can set your NIC to force itself to a particular speed and duplex by using the -F switch with any of the following options: 100baseTx-FD. 10baseT-FD. with the NIC selecting 100Mbps. link ok [root@bigboy tmp]# By using the verbose mode -v switch you can get much more information. link ok product info: vendor 00:10:18. A list of supported modes is also provided at the top of the output: [root@bigboy tmp]# ethtool eth0 Settings for eth0: . full duplex.” Using ethtool The ethtool command is slated to be the replacement for mii-tool in the near future and tends to be supported by newer NIC cards.qxd 2/25/05 9:56 AM Page 50 50 Linux Networking Chapter 3 [root@bigboy tmp]# mii-tool SIOCGMIIPHY on ‘eth0’ failed: Operation not supported eth1: 100 Mbit.local file to let it be run at the very end of the booting process or by creating your own startup script if you need it set earlier. full duplex mode (FD): [root@bigboy tmp]# mii-tool -v eth1: negotiated 100baseTx-FD. Remember that you could lose all network connectivity to your server if you force your NIC to a particular speed/duplex that doesn’t match that of your switch: [root@bigboy tmp]# mii-tool -F 100baseTx-FD eth0 Unfortunately there is no way to set this on reboot permanently except by placing the command in the /etc/rc. or 10baseT-HD.

1. which provides a quicker summary: [root@bigboy tmp]# ethtool . [root@bigboy tmp]# ethtool -s eth1 speed 100 duplex full autoneg off [root@bigboy tmp]# The Linux man pages give more details on other ethtool options.255..255. . but you can get a quick guide by just entering the ethtool command alone..100 NETMASK=255.04Harrison_ch03.. . ethtool settings can be permanently set as part of the interface’s configuration script with the ETHTOOL_OPTS variable.. These settings can also be changed from the command line using the -s switch followed by the interface name and its desired configuration parameters. full duplex with no chance for autonegotiation on the next reboot: # # File: /etc/sysconfig/network-script/ifcfg-eth0 # DEVICE=eth0 IPADDR=192.qxd 2/25/05 9:56 AM Page 51 Changing NIC Speed and Duplex 51 Supported ports: [ TP MII ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full Advertised auto-negotiation: No Speed: 100Mb/s Duplex: Full Port: MII PHYAD: 1 Transceiver: internal Auto-negotiation: off Supports Wake-on: g Wake-on: g Current message level: 0x00000007 (7) Link detected: yes [root@bigboy tmp]# Setting Your NIC’s Speed Parameters with ethtool Unlike mii-tool.0 BOOTPROTO=static ONBOOT=yes ETHTOOL_OPTS = “speed 100 duplex full autoneg off” You can test the application of these parameters by shutting down the interface and activating it again with the ifup and ifdown commands.168. In our next example. the settings will be set to 100Mbps.

Here is how it’s done: [root@bigboy tmp] sysctl -p sysctl -p net.ipv4. Configuring IP Forwarding For your Linux server to become a router.ip_forward=0 After: # Enables packet forwarding net. Details on how to configure Linux firewall security are covered in Chapter 14.conf..ip_forward and can be found in the file /etc/sysctl.ipv4. The Linux kernel configuration parameter to activate this is named net.ip_forward=1 This enables packet forwarding only when you reboot. you have to enable packet forwarding..04Harrison_ch03. . Before: # Disables packet forwarding net. but when budgets are tight you might seriously want to consider modifying an existing Linux server to do the job. To activate the feature immediately you have to force Linux to read the /etc/sysctl. [root@bigboy tmp]# HOW TO CONVERT YOUR LINUX SERVER INTO A ROUTER Router/firewall appliances that provide basic Internet connectivity for a small office or home network are becoming more affordable every day. but you need to understand how to activate routing through the firewall before it can become a functioning networking device.ip_forward = 1 net.default.ipv4..conf.qxd 2/25/05 9:56 AM Page 52 52 Linux Networking Chapter 3 ethtool -s DEVNAME \ [ speed 10|100|1000 ] \ [ duplex half|full ] \ [ port tp|aui|bnc|mii|fibre ] \ .ipv4.rp_filter = 1 .conf file with the sysctl command using the -p switch. at which time Linux will create a file in one of the subdirectories of the special RAM memory-based /proc filesystem. Remove the # from the line related to packet forwarding..ipv4. In simple terms packet forwarding enables packets to flow through the Linux box from one network to another.

This feature is called proxy ARP.” for more information on adjusting kernel parameters. If one of the routers has an incorrect routing table entry for the remote network. Note It is for this and other reasons that it is generally not a good idea to configure proxy ARP on a router. “Miscellaneous Linux Topics.qxd 2/25/05 9:56 AM Page 53 How to Convert Your Linux Server into a Router 53 kernel. The server then sends an ARP request for the MAC address that matches the router’s IP address. If the same server needs to send a packet to another device on a remote network the process is different. there is the risk that traffic to the remote server will occasionally get lost. The firewall typically isn’t configured with an IP address on the LAN and appears to be an intelligent cable capable of selectively blocking packets. Some routers can be configured to answer these types of ARP requests for remote networks. It then sends the packet to the router using the router’s MAC address and a destination IP address of the remote server. In other words you can lose routing control. it sends out an ARP request to the network asking for the MAC address of the other device. This directory then has subdirectories corresponding to each functioning NIC card on your server. The server first takes a look at its routing table to find out the IP address of the best router on its network that will be able to relay the packet to the destination. In this scenario there is the possibility that either one will answer the local server’s ARP request for the MAC address of the remote server. These devices are typically inserted as part of a daisy chain connecting multiple network switches on the same LAN while protecting one section of a LAN from traffic originating on another section.sysrq = 0 kernel. It is also good to always configure a default gateway on your server and use separate routing entries via other routers for all networks your default gateway may not know about. Each subdirectory .04Harrison_ch03. One of the most common problems occurs if two routers are on the network configured for proxy ARP.core_uses_pid = 1 [root@bigboy tmp]# Please refer to Appendix I. the server then sends out an ARP request for the MAC address of the remote server. Configuring Proxy ARP If a server needs to send a packet to another device on the same network. There are some disadvantages with this. Some types of bridging mode firewalls need to have proxy ARP enabled to operate properly. If you need to enable proxy ARP on a Linux server the /proc filesystem comes into play again. Proxy ARP is handled by files in the /proc/sys/net/ipv4/conf/ directory. If there is no suitable router on its network.

first for all interfaces and then for interfaces eth0 and wlan0: # # File: /etc/sysctl. Unfortunately.conf # # Enables Proxy ARP on all interfaces net/ipv4/conf/all/proxy_arp = 1 # Enables Proxy ARP on interfaces eth1 and wlan0 net/ipv4/conf/eth1/proxy_arp = 1 net/ipv4/conf/wlan0/proxy_arp = 1 You can then activate these settings with the sysctl command: [root@bigboy tmp] sysctl -p CONFIGURING YOUR /ETC/HOSTS FILE The /etc/hosts file is just a list of IP addresses and their corresponding server names.101 smallfry In this example. and in such cases it may be easier to add a quick /etc/hosts file entry until the centralized change can be made: 192. you also have to update the file. if the value is 1.1. telnet. server smallfry has an IP address of 192. but can become laborious if it has to be done companywide. If the value within this file is 0. The next example activates proxy ARP.1.101.101 using ping.101) 56(84) bytes of data.1. You can access 192. if the IP address for that host changes. For ease of management. DNS won’t be queried at all. Here is an example using the ping command to see whether smallfry is alive and well on the network: [root@bigboy tmp]# ping smallfry PING zero (192.04Harrison_ch03. or any other network-aware program by referring to it as smallfry. it is often easiest to limit entries in this file to just the loopback interface and also the server’s own hostname.168.168.conf file mentioned in Appendix II to activate or disable proxy ARP.qxd 2/25/05 9:56 AM Page 54 54 Linux Networking Chapter 3 then has a file called proxy_arp. This may not be much of a concern for a single server.168.1. You can use the /etc/sysctl. . If the name is found with a corresponding IP address. it is enabled. proxy ARP on the interface is disabled.168. and use a centralized DNS server to handle most of the rest. Sometimes you may not be the one managing the DNS server. Your server will typically check this file before referencing DNS.

1 bigboy.1. This can cause confusion.0. This is usually mapped to the name localhost.101 smallfry tiny littleguy # (Wrong) # (Wrong) # (Wrong) The loopback Interface’s localhost Entry Usually the first entry in /etc/hosts defines the IP address of the server’s virtual loopback interface.1. because the other server also feels it is localhost. 0% packet loss. 2 received. Here we have set it up so that smallfry can also be accessed using the names tiny and littleguy: 192.localdomain. pipe 2 [root@bigboy tmp]# You can also add aliases to the end of the line.101): icmp_seq=0 ttl=64 time=0.168. this first entry after the 127.localdomain when communicating with another server on the network.034/0.smallfry ping statistics --2 packets transmitted.101 192.qxd 2/25/05 9:56 AM Page 55 Configuring Your /etc/hosts File 55 64 bytes from smallfry (192.101 smallfry tiny littleguy You should never have an IP address more than once in this file because Linux will use only the values in the first entry it finds: 192.my-web-site.197 ms 64 bytes from smallfry (192.1 needs to be the fully qualified domain name (FQDN) of the server— for example.0.168.0.168.0.101 192. which enable you to refer to the server using other names.04Harrison_ch03.047 ms --.074 ms.0.com: 127.168.1 and the localhost entries like this: 127.0.1.localdomain localhost When the server is connected to the Internet. By default. .1 bigboy localhost.0.0. Fedora inserts the hostname of the server between the 127. bigboy.1.101): icmp_seq=1 ttl=64 time=0.mysite.168.1.197/0.092/0.org localhost.localdomain localhost Some programs such as Sendmail are very sensitive to this and if they detect what they feel is an incorrect FQDN they will default to using the name localhost.1. time 2017ms rtt min/avg/max/mdev = 0.localdomain (the universal name used when a server refers to itself) and localhost (the shortened alias name).168.

CONCLUSION As you can imagine.localdomain entry mapping to 127. . The skills you develop to identify and rectify these issues could prove to be invaluable to your company and career. Good systems administrators know the tools needed to be able to identify the probable causes of these types of problem. There always things that can go wrong that may be totally out of your control.04Harrison_ch03.0. The next two chapters show you how to test your network and Linux server applications confidently when things appear to go wrong.1 for Linux to work properly and securely. which enables them to know the type of help they need to fix it. configuring Linux networking is just a first step in providing Internet access to your server.qxd 2/25/05 9:56 AM Page 56 56 Linux Networking Chapter 3 Note You must always have a localhost and localhost.0.

05Harrison_ch04.qxd 2/25/05 9:56 AM Page 57 C H A P T E R 4 Simple Network Troubleshooting IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Doing Basic Cable and Link Tests Testing Your NIC How to See MAC Addresses Using ping to Test Network Connectivity Using telnet to Test Network Connectivity Testing Web Sites with the curl and wget Utilities The netstat Command The Linux iptables Firewall Using traceroute to Test Connectivity Using MTR to Detect Network Congestion Viewing Packet Flows with tcpdump Viewing Packet Flows with tethereal Basic DNS Troubleshooting Using nmap Determining the Source of an Attack Who Has Used My System? Conclusion You will eventually find yourself trying to fix a network related problem. These symptoms can be caused by: 57 . which usually appears in one of two forms. The first is slow response times from the remote server. and the second is a complete lack of connectivity.

More sophisticated models on the market will be able to tell you the approximate location of a cable break and whether an Ethernet cable is too long to be used.” there are two types of Ethernet cables: crossover and straight-through. ☞ The cables aren’t plugged in properly. Additional sources of disconnections are: ☞ Power failures ☞ The remote server or an application on the remote server being shut down. investment in a battery-operated cable tester for basic connectivity testing is invaluable. We discuss how to isolate these problems and more in the following sections.qxd 2/25/05 9:56 AM Page 58 58 Simple Network Troubleshooting Chapter 4 Sources of Network Slowness ☞ ☞ ☞ ☞ ☞ ☞ ☞ NIC duplex and speed incompatibilities Network congestion Poor routing Bad cabling Electrical interference An overloaded server at the remote end of the connection Misconfigured DNS (Covered in Chapter 18. Always make sure you are using the correct type. If you have an extensive network. “Introduction to Networking. This indicates that the connection between your server and the switch/router is functioning correctly. In most cases a lack of link is due to the wrong cable type being used. ☞ The switch or router to which the server is connected is powered down. As described in Chapter 2. Other sources of link failure include: ☞ The cables are bad.” and Chapter 19. “Dynamic DNS”) Sources of a Lack of Connectivity All sources of slowness can become so severe that connectivity is lost. .05Harrison_ch04. “Configuring DNS. DOING BASIC CABLE AND LINK TESTS Your server won’t be able to communicate with any other device on your network unless the NIC’s link light is on.

7 Kb) TX bytes:166377 (162.05Harrison_ch04.119. whether they are functional or not.qxd 2/25/05 9:56 AM Page 59 Testing Your NIC 59 TESTING YOUR NIC It is always a good practice in troubleshooting to be versed in monitoring the status of your NIC from the command line.4 Kb) Interrupt:10 Memory:c88b5000-c88b6000 Active Interface wlan0 Link encap:Ethernet HWaddr 00:06:25:09:6A:D7 inet addr:216. Shut Down Interface wlan0 Link encap:Ethernet HWaddr 00:06:25:09:6A:D7 BROADCAST MULTICAST MTU:1500 Metric:1 RX packets:2924 errors:0 dropped:0 overruns:0 frame:0 TX packets:2287 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:180948 (176.119. Interfaces that are shut down by the systems administrator or are nonfunctional will not show an IP address line and the word UP will not show in the second line of the output. Viewing Your Activated Interfaces The ifconfig command without any arguments gives you all the active interfaces on your system.6 Kb) Interrupt:10 Memory:c88b5000-c88b6000 . We’ll soon discuss how to determine the link status using commands. This can be seen in the next examples. Interfaces will not appear if they are shut down: [root@bigboy tmp]# ifconfig Note Interfaces will appear if they are activated but have no link. The following sections introduce a few commands that will be useful.255 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:2924 errors:0 dropped:0 overruns:0 frame:0 TX packets:2295 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:180948 (176.243 Bcast:216.7 Kb) TX bytes:166521 (162. Viewing All Interfaces The ifconfig -a command provides all the network interfaces.10.10.

mii-tool gives a very brief report. link ok product info: Intel 82555 rev 4 basic mode: 100 Mbit. and this can be useful in troubleshooting speed and duplex issues. Use it with the -v switch because it provides more information on the supported autonegotiation speeds of the NIC. especially because mii-tool will be soon deprecated in Linux. full duplex. In both of the following examples.qxd 2/25/05 9:56 AM Page 60 60 Simple Network Troubleshooting Chapter 4 Testing Link Status from the Command Line Both the mii-tool and ethtool commands provide reports on the link status and duplex settings for supported NICs. and the link is ok. the NICs are operating at 100Mbps. The ethtool command provides much more information than mii-tool and should be your command of choice.05Harrison_ch04. full duplex basic status: link ok capabilities: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control link partner: 100baseTx-HD [root@bigboy tmp]# Link Status Output from ethtool [root@bigboy tmp]# ethtool eth0 Settings for eth0: Supported ports: [ TP MII ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full Advertised auto-negotiation: No Speed: 100Mb/s Duplex: Full Port: MII PHYAD: 1 Transceiver: internal Auto-negotiation: off Supports Wake-on: g Wake-on: g Current message level: 0x00000007 (7) Link detected: yes [root@bigboy tmp]# . When used without any switches. Link Status Output from mii-tool [root@bigboy tmp]# mii-tool -v eth0: 100 Mbit. full duplex.

qxd 2/25/05 9:56 AM Page 61 Testing Your NIC 61 Viewing NIC Errors Errors are a common symptom of slow connectivity due to poor configuration or excessive bandwidth utilization.5% can result in noticeable sluggishness.7 Kb) TX bytes:166377 (162. and frame errors.05Harrison_ch04. They should always be corrected whenever possible.4 Kb) Interrupt:10 Memory:c88b5000-c88b6000 ethtool Error Output The ethtool command can provide a much more detailed report when used with the -S switch: [root@probe-001 root]# ethtool -S eth0 NIC statistics: rx_packets: 1669993 tx_packets: 627631 rx_bytes: 361714034 tx_bytes: 88228145 rx_errors: 0 tx_errors: 0 rx_dropped: 0 tx_dropped: 0 multicast: 0 collisions: 0 rx_length_errors: 0 rx_over_errors: 0 rx_crc_errors: 0 rx_frame_errors: 0 rx_fifo_errors: 0 rx_missed_errors: 0 tx_aborted_errors: 0 tx_carrier_errors: 0 tx_fifo_errors: 0 tx_heartbeat_errors: 0 tx_window_errors: 0 tx_deferred: 0 tx_single_collisions: 0 tx_multi_collisions: 0 tx_flow_control_pause: 0 rx_flow_control_pause: 0 rx_flow_control_unsupported: 0 tx_tco_packets: 0 rx_tco_packets: 0 [root@probe-001 root]# . Ifconfig Error Output The ifconfig command also shows the number of overrun. wlan0 Link encap:Ethernet HWaddr 00:06:25:09:6A:D7 BROADCAST MULTICAST MTU:1500 Metric:1 RX packets:2924 errors:0 dropped:0 overruns:0 frame:0 TX packets:2287 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:180948 (176. dropped packet. Error rates in excess of 0. carrier.

usually is an indication of electrical noise. and that the connectors are securely fastened. ☞ Multiple Collisions: The NIC had to attempt multiple times before successfully sending the frame due to collisions. ☞ There might be bad network cabling. ☞ CRC Errors: Frames were sent but were corrupted in transit. . ☞ A NIC might be disabled or the remote server might be shut down.qxd 2/25/05 9:56 AM Page 62 62 Simple Network Troubleshooting Chapter 4 Possible Causes of Ethernet Errors The following are possible causes of Ethernet errors: ☞ Collisions: The NIC card detects itself and another server on the LAN attempting data transmissions at the same time. Higher error rates are likely to be caused by faulty NIC cards or poorly terminated cables. This is usually a sign of excessive traffic.1% of all frames sent. Lack of communication at this level may mean: ☞ Either server might be disconnected from the network. HOW TO SEE MAC ADDRESSES There are times when you lose connectivity with another server that is directly connected to your local network. ☞ Length Errors: The received frame length was less than or exceeded the Ethernet standard. Check for faulty cabling or faulty interfaces on the NIC and networking equipment. This is usually the result of collisions or a bad Ethernet device. ☞ Single Collisions: The Ethernet frame went through after only one collision.05Harrison_ch04. Taking a look at the ARP table of the server from which you are troubleshooting will help determine whether the remote server’s NIC is responding to any type of traffic from your Linux box. Make sure that you are using the correct type of cable. Collisions can be expected as a normal part of Ethernet operation and are typically below 0. The presence of CRC errors. ☞ Carrier Errors: Errors are caused by the NIC losing its link connection to the hub or switch. but not many collisions. that the cabling is undamaged. This is most frequently due to incompatible duplex settings. ☞ FIFO and Overrun Errors: The number of times that the NIC was blocked from transferring data from the network to its memory buffers because of the speed limitations of the hardware. ☞ Frame Errors: An incorrect CRC and a noninteger number of bytes are received.

99 tied to the NIC hardware MAC address of 00:06:25:09:6A:B5: ☞ The arp -a command will show you the MAC addresses in your server’s ARP table and all the other servers on the directly connected network.168. . Here we see we have some form of connectivity with the router at address 192.1.255.255.168.4 Mb) TX bytes:43209032 (41. you can see the MAC address.99 Bcast:192.1) at 00:09:E8:9C:FD:AB [ether] on wlan0 ? (192.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:47379 errors:0 dropped:0 overruns:0 frame:0 TX packets:107900 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:4676853 (4.1.101) at 00:06:25:09:6A:D7 [ether] on wlan0 [root@bigboy tmp]# ☞ Make sure the IP addresses listed in the ARP table match those of servers expected to be on your network. Here is a description of the commands you may use to determine ARP values: ☞ The ifconfig -a command shows you both the NIC’s MAC address and the associated IP addresses of the server that you are currently logged in to: [root@bigboy tmp]# ifconfig -a wlan0 Link encap:Ethernet HWaddr 00:06:25:09:6A:B5 inet addr:192.255.1: [root@bigboy tmp]# arp -a bigboypix (192. ☞ You should also check the ARP table of the remote server to see whether it is populated with acceptable values. If they don’t.255 Mask:255.1.1.1.1.168.100 and 192.1.1.168.168.1.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 Interrupt:11 Memory:c887a000-c887b000 [root@bigboy tmp]# Here you can see the wlan0 interface has two IP addresses 192. the server is running the correct software.168. Typically in this case.168.qxd 2/25/05 9:56 AM Page 63 How to See MAC Addresses 63 ☞ The remote server might be running firewall software such as iptables or the Windows XP built-in firewall. your server may be plugged into the wrong switch or router port.255.168.255 Mask:255.100 Bcast:192. but the desired communication doesn’t appear to be occurring to the client on the same network.2 Mb) Interrupt:11 Memory:c887a000-c887b000 wlan0:0 Link encap:Ethernet HWaddr 00:06:25:09:6A:B5 inet addr:192.168.05Harrison_ch04.

07 time=4. See the appendix for a full listing of them. 0% loss.31 ms ms ms ms --.101) from of data.95 time=7.168. This message is caused by your router or server knowing that the target IP address is part of a valid network.168. You have incorrect routing. time 3026ms rtt min/avg/max/mdev = 3. it becomes a very handy tool.168.100 : 56(84) bytes ttl=128 ttl=128 ttl=128 ttl=128 time=3.1. Try pinging the gateways to make sure they are responsive.100: [root@smallfry tmp]# ping 192.168.101 ping statistics --4 packets transmitted.168.101: icmp_seq=2 64 bytes from 192.05Harrison_ch04. 64 bytes from 192.242 ms [root@smallfry tmp]# You may get a Destination Host Unreachable message.101: icmp_seq=1 64 bytes from 192.1.101: icmp_seq=4 192. The Linux ping command sends continuous pings. ping sends ICMP echo packets that request a corresponding ICMP echo-reply response from the device at the target address.192.1. A firewall or router along the network path is blocking ICMP traffic.168. 4 received.101: icmp_seq=3 64 bytes from 192.1. There are a variety of ICMP response codes that can help in further troubleshooting.1.1. it is always a good practice to force a response from it. your traffic won’t be able to leave the LAN. Use traceroute to ensure you’re taking the correct path. but is getting no response from the target server. A classic symptom of bad routes on a server is the ability to ping servers only on your local network and nowhere else. If you can’t get to the first network hop.101 (192.1. One of the most common methods used to test connectivity across multiple networks is the ping command. Check the routes and subnet masks on both the local and remote servers and all routers in between. once a second.1.168.072/1.168.168. until stopped with a Ctrl-C.1. ☞ You may have lost connectivity to the gateway router on the network. A lack of response could be due to: ☞ ☞ ☞ ☞ A server with that IP address doesn’t exist.qxd 2/25/05 9:56 AM Page 64 64 Simple Network Troubleshooting Chapter 4 USING PING TO TEST NETWORK CONNECTIVITY Whether or not your troublesome server is connected to your local network.101 PING 192. The server has been configured not to respond to pings.1.948/7. .168. Because most servers will respond to a ping query. There are a number of reasons for this. Here is an example of a successful ping to the server bigboy at 192.46 time=4. ☞ Either the source or destination device has an incorrect IP address or subnet mask.950/4.

105 PING 192. Here is an example of testing server 192.100 icmp_seq=6 Destination Host Unreachable --.168. ☞ You might have the incorrect type of cable connecting your Linux box to the network.102 22 When using telnet troubleshooting. HTTPS uses port 443.1.1.168.168.100 icmp_seq=1 Destination Host Unreachable From 192. +6 errors. telnet will try to connect on TCP port 23.05Harrison_ch04.100 icmp_seq=3 Destination Host Unreachable From 192.1.1.1.168.168. By default.100 icmp_seq=4 Destination Host Unreachable From 192.1.1.102 on the TCP port 22 reserved for SSH: [root@bigboy tmp]# telnet 192. but you can specify other TCP ports by typing them in after the target IP address. 0 received. you may verify this with the mii-tool command. ☞ In the case of a wireless network.1.168.168. .100 icmp_seq=2 Destination Host Unreachable From 192.1.1. time 7021ms.168.100 icmp_seq=5 Destination Host Unreachable From 192. There are two basic types: straight through and crossover.168.105 ping statistics --8 packets transmitted.105 (192. which triggers the message. 100% loss.1.168. HTTP uses TCP port 80.105) from 192. ☞ Your NIC might not have the correct duplex settings. your SSID or encryption keys might be incorrect If you are trying to ping a host on a remote network: ☞ The network device doesn’t have a route in its routing table to the destination network and sends an ICMP reply type 3.1.1.168. here are some useful guidelines to follow that will help isolate the source of the problem: ☞ Test connectivity from the remote PC or server. pipe 3 [root@smallfry tmp]# USING TELNET TO TEST NETWORK CONNECTIVITY An easy way to tell if a remote server is listening on a specific TCP port is to use the telnet command.qxd 2/25/05 9:56 AM Page 65 Using telnet to Test Network Connectivity 65 If you are trying to ping a host on a directly connected network: ☞ The server might be down or disconnected for the network.168.192. The resulting message may be Destination Host Unreachable or Destination Network Unreachable: [root@smallfry tmp]# ping 192. From 192.168.100 : 56(84) bytes of data.

99-OpenSSH_3.1. but connectivity to desired TCP ports on the NIC interface may be blocked.168. If the server is running a firewall package. Further discussion of the Linux iptables package is covered in a later section. LINUX TELNET TROUBLESHOOTING telnet The following sections illustrate the use of Linux box.1.1. all loopback connectivity could be allowed. troubleshooting from a Note Always remember that many Linux servers have the iptables firewall package installed by default. Connected to 192. such as the Linux iptables software.102. .05Harrison_ch04.102. You can use the /etc/init.. iptables may be safely turned off. This helps eliminate the influence of any firewalls protecting the entire network from outside. In some cases where the network is already protected by a firewall.. [root@ bigboy tmp]# To break out of the connection you have to press the Ctrl and ] keys simultaneously.168.1.102 on the SSH port (TCP 22): [root@bigboy tmp]# telnet 192. This is often the cause of many connectivity problems and the firewall rules should be correctly updated. Escape character is ‘^]’. SSH-1.102 22 Trying 192.168. not the usual Ctrl-C.4p1 ^] telnet> quit Connection closed.qxd 2/25/05 9:56 AM Page 66 66 Simple Network Troubleshooting Chapter 4 ☞ Test connectivity on the server itself. a successful telnet connection is always greeted by a Connected to message like this one when trying to test connectivity to server 192.d/iptables status command on the target server to determine whether iptables is running.168. Successful Connection With Linux. ☞ Test connectivity from another server on the same network as the target server. Try making the connection to the loopback address as well as the NIC IP address.

. These symptoms can be caused by one of the following reasons: ☞ The remote server doesn’t exist on the destination network.168. This is also known as hanging. It could be turned off. In some cases. This is usually caused by there being correct network connectivity but a poorly configured application.100.100.12: Connection timed out [root@bigboy tmp]# TELNET TROUBLESHOOTING USING WINDOWS Sometimes you have to troubleshoot Linux servers from a Windows PC..168. Connection Refused Messages You will get a Connection refused message for one of the following reasons: ☞ The application you are trying to test hasn’t been started on the remote server. Here is some sample output: [root@bigboy tmp]# telnet 192. but the results are different.12 22 Trying 216.168.100: Connection refused [root@bigboy tmp]# telnet Timeout or Hanging The telnet command will abort the attempted connection after waiting a predetermined time for a response. yet the application doesn’t appear to work. Go to the command line and type the same telnet command as you would in Linux.10. . This is called a timeout. telnet: connect to address 192.12.qxd 2/25/05 9:56 AM Page 67 telnet Troubleshooting Using Windows 67 Note In many cases you can successfully connect on the remote server on the desired TCP port. causing it to time out instead of being quickly refused.1.1..100 22 Trying 192.. [root@bigboy tmp]# telnet 216.100. ☞ There is a firewall blocking and rejecting the connection attempt.100.05Harrison_ch04. telnet: connect to address 216. The telnet commands are the same. telnet won’t abort.10.1. ☞ A firewall could be blocking and not rejecting the connection attempt. but will just wait indefinitely.10.

but to slowness in the configuration of the Web server or any supporting application or database servers it may use to generate the Web page. A much better combination of tools is to use telnet to test your site’s TCP port 80 response time in conjunction with data from the Linux curl and wget HTTP utilities.1. Connect Failed Messages The Connect failed messages are the equivalent of the Linux Connection refused messages explained earlier and are caused by the same reasons: C:\>telnet 172. your command prompt screen will go blank. usually point not to a network issue.qxd 2/25/05 9:56 AM Page 68 68 Simple Network Troubleshooting Chapter 4 Screen Goes Blank—Successful Connection If there is connectivity.102 256 Connecting To 172. C:\>telnet 216.100.. Using curl The curl utility acts like a text-based Web browser in which you can select to see either the header or complete body of a Web page’s HTML code displayed on your screen.100.1. Rapid TCP response times. Using the Ctrl-C key sequence enables you to exit the telnet attempt.102..Could not open connection to the host. the session will appear to hang or time out.10.05Harrison_ch04. on port 256: Connect failed C:\> telnet Timeout or Hanging As explained previously. but slow curl and wget response times. This is usually caused by the target server being turned off or by a firewall blocking the connection. making troubleshooting difficult. .12 22 Connecting To 216.12. TESTING WEB SITES WITH THE CURL AND WGET UTILITIES Testing a Web site’s performance using a Web browser alone is sometimes insufficient to get a good idea of the source of slow Web server performance.16. Many useful HTTP error codes are often not displayed by browsers.16.10.. if there is no connectivity..

you view not only the HTML content of the Web site’s index page in your local directory. 19 Oct 2004 05:11:22 GMT Server: Apache/2. Please refer to Chapter 20.com.com/ => `index.linuxhomenetworking..36 KB/s) .com[65... Connecting to www. This can be very helpful in providing a simple way to obtain snapshots of your server’s performance: [root@zippy tmp]# wget -N www. By not using the -I flag you will see all the Web page’s HTML code displayed on the screen.linuxhomenetworking. HTTP request sent.html’ saved [122150] [root@zippy tmp]# ..`index. connected. By not using recursion and activating the time stamping feature (the -N switch). done.34]:80. connected.time-stamps turned off.71. HTTP request sent. including the entire directory structure of the Web site.115.51 (Fedora) Accept-Ranges: bytes Vary: Accept-Encoding..com/ => `index.html’ Connecting to www. 200 OK Length: unspecified [text/html] [ <=> 279.qxd 2/25/05 9:56 AM Page 69 Testing Web Sites with the curl and wget Utilities 69 A good start is to use the curl command with the -I flag to view just the Web page’s header and HTTP status code. awaiting response. Using wget You can use wget to recursively download a Web site’s pages. “The Apache Web Server. awaiting response. 200 OK Length: unspecified [text/html] Last-modified header missing -.linuxhomenetworking.34]:80.” for a more complete listing of possibilities.1 200 OK Date: Tue. Either method can provide a good idea of your server’s performance: [root@ bigboy tmp]# curl -I www. but also the download speed... --23:07:22-.linuxhomenetworking.User-Agent Connection: close Content-Type: text/html.html’ Resolving www..com[65.. and precise start and stop times for the download.linuxhomenetworking.http://www.0.linuxhomenetworking.http://www.150 23:07:22 (279.115. charset=UTF-8 [root@bigboy tmp]# In this case the Web server appears to be working correctly because it returns a 200 OK code.com HTTP/1.71.36K/s ] 122..linuxhomenetworking.com --23:07:22-. to a local directory. file size.05Harrison_ch04.

255. The traceroute command.0 255. which in this case shows 14 connections: [root@bigboy tmp]# netstat -an | grep tcp | egrep -i \ ‘established|time_wait’ | wc -l 14 [root@bigboy tmp]# The netstat -nr command can also be used to view your routing table. . with the number of matches being counted by the wc command..68.0 255.135 255.193 0.0.0 172.16.131 255.255.160.255. If the first hop is incorrect.224 172.0.0.0.193 255.0:* tcp 0 0 :::80 :::* .qxd 2/25/05 9:56 AM Page 70 70 Simple Network Troubleshooting Chapter 4 THE NETSTAT COMMAND Like curl and wget.135 255.69.16.16.16.0. including all the active network connections to and from your server.67.255.193 255. netstat can be very useful in helping determine the source of problems.0 172.16.0 127.16.11.69. then your routing table needs to be examined more carefully.0..0 172..255.0 172.0 [root@bigboy tmp]# Flags MSS Window irtt UG UG UG UG U U UG UG U UG 40 40 40 40 40 40 40 40 40 40 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 eth1 eth1 eth1 eth0 eth1 eth0 eth0 eth0 lo eth1 .0.69.64 172.192 172.16.0.0.69.128 172.16.255.0 0. It is therefore a good idea to focus on these types of short-lived connections. Using netstat with the -an option lists all the TCP ports on which your Linux server is listening.0..255. [root@bigboy tmp]# State LISTEN LISTEN Most TCP connections create permanent connections.193 255.0.0.69.255..255.16.128 0.0.255.16.224 172.192 0. which I’ll discuss later.255.0 0.16.0.1:25 0. This can be very helpful in determining whether slowness is due to high traffic volumes: [root@bigboy tmp]# netstat -an Active Internet connections (servers and established) Proto Recv-Q Send-Q Local Address Foreign Address tcp 0 0 127.224 172.0 172. You can determine the number of established and time_wait TCP connections on your server by using the netstat command filtered by the grep and egrep commands.0.0.255.. HTTP is different because the connections are shut down on their own after a predefined inactive timeout or time_wait period on the Web server.68.32 172. . can then be used to verify that your routing table is correct by displaying the path a packet takes to get to a remote destination.0.0 255.0.16.0. It is always good to ensure that your routes are correct and that you can ping all the gateways in your routing table.67.67.96 172.240. [root@bigboy tmp]# netstat -nr Kernel IP routing table Destination Gateway Genmask Iface 172.255.16.16.224 172.05Harrison_ch04.0.67.0.67.

you may want to temporarily disable iptables using the same /etc/init.” USING TRACEROUTE TO TEST CONNECTIVITY Another tool for network troubleshooting is the traceroute command.d/ iptables script with the status qualifier. but also sends an ICMP time exceeded message back to the source. You can read more about configuring iptables in Chapter 14. [root@zero root]# How to Stop iptables If your Linux box is already protected by a firewall.05Harrison_ch04. The first router on the route recognizes that the TTL has already been exceeded and discards or drops the packet. “Linux Firewalls Using iptables. The traceroute program records the IP address of the router that sent the message and knows that .d/iptables script with the stop qualifier: [root@bigboy tmp]# service iptables stop Flushing firewall rules: [ OK ] Setting chains to policy ACCEPT: filter [ Unloading iptables modules: [ OK ] [root@bigboy tmp]# OK ] How to Configure iptables Rules Stopping iptables may not be a good permanent solution. If it isn’t running you’ll get a very short listing of the firewall rules.qxd 2/25/05 2:34 PM Page 71 Using traceroute to Test Connectivity 71 THE LINUX IPTABLES FIREWALL An unexpected source of server connectivity issues for brand new servers is frequently the iptables firewall. Here is some sample output: [root@zero root]# service iptables status Firewall is stopped. The traceroute command works by sending a UDP packet destined to the target with a TTL of 0. It gives a listing of all the router hops between your server and the target server. This is installed by default under Fedora and Red Hat and usually allows only a limited range of traffic. Determining Whether iptables Is Running You can easily test whether iptables is running by using the /etc/init. especially if your network isn’t protected by a firewall. This helps you verify that routing over the networks in between is correct.

net (67.another-isp-provider.120.164 ms 6 sl-gw11-sj-3-0.sntc01.064 ms 13.111 ms 2 dist3-vlan50.20. This continues around and around until the final destination is reached. 30 hops max. these are .net (144.20.net (63.net (144.158 traceroute to 144. drops the packet and replies with an ICMP time exceeded message.sntc01. The first hop sees nothing wrong with the packet.my-isp-provider.67) 13. traceroute now knows the IP address of the second router.my-isp-provider.185 ms 15.745 ms 4 bb2-p11-0.my-isp-provider.887 ms 13.133) 16.161.20. with a TTL of 1.sntc01.854 ms 13.20. it is less likely to be the source of your problems.158).221. and forwards the packet to the second hop on the path.05Harrison_ch04.my-isp-provider. decrements the TTL to 0 as expected. The traceroute program tries again.232.618 ms 15.260 ms 15.161.17) 12.49) 14.785 ms 17.my-isp-provider.net (63.035 ms 13.423 ms 8 sl-bb23-ana-6-0. 38 byte packets 1 adsl-67-120-221-110.another-isp-provider. Note In Linux the traceroute command is traceroute.snfc21.20.net (144.232.dsl.232.net (64.203.another-isp-provider.53) 15.306 ms 26.110) 14.158) 27. In Windows it is tracert. Router 2 sees the TTL of 0.158.246) 16.482 ms 26. Sample traceroute Output Here is a sample output for a query to 144. If a device responds.663 ms 5 bb1-p14-0.snfc21.232.qxd 2/25/05 9:56 AM Page 72 72 Simple Network Troubleshooting Chapter 4 that is the first hop on the path to the final destination.net (64.018 ms 12.1.35.443 ms 16.232.35.146 ms 3 bb1-g1-0.897 ms 15.228. Note You receive traceroute responses only from functioning devices. traceroute can give.124.124.408 ms 14.857 ms 15.44.158 (144.189 ms 7 sl-bb25-sj-6-1.3.487 ms [root@bigboy tmp]# Possible traceroute Messages There are a number of possible message codes listed in Table 4.279 ms 15.232.203. Notice that all the hop times are under 50 milliseconds (ms) which is acceptable: [root@bigboy tmp]# traceroute -I 144.

158 traceroute to 144. Source route failed.another-isp-provider. Failure may be due to a router security setting.my-isp-provider.161. !N.my-isp-provider.116 ms 7 sl-bb25-sj-6-1. Could be caused by the following: ☞ A router on the path not sending back the ICMP time exceeded messages ☞ A router or firewall in the path blocking the ICMP time exceeded messages ☞ The target IP address not responding Host.49) 16. or protocol unreachable.53) 16.my-isp-provider.627 ms 3 bb1-g1-0.203.521 ms 12.3.sntc01.728 ms 5 bb1-p14-0.232.971 ms 14. or !P !X or !A !S traceroute Time Exceeded False Alarms If there is no response within a 5-second timeout interval.1 traceroute Return Code Symbols Traceroute Symbol *** Description Expected 5-second response time exceeded.20.my-isp-provider.158).000 ms 14.510 ms 15.snfc21. but will allow ICMP packets.120. 38 byte packets 1 adsl-67-120-221-110.sntc01.another-isp-provider.133) 16.228.snfc21.158 (144.161.274 ms 15. In this case the * * * status messages disappear: .net (64.net (63.net (63.67) 12.35.203.232.110) 14. 30 hops max.860 ms 13. Source routing attempts to force traceroute to use a certain path.991 ms 15.469 ms 18.05Harrison_ch04.20.926 ms 8 * * * 9 * * * [root@bigboy tmp]# Some devices will prevent traceroute packets directed at their interfaces. as seen in the following example: [root@bigboy tmp]# traceroute 144. Using traceroute with an -I flag forces traceroute to use ICMP packets that may go through.510 ms 17.019 ms 16.net (67.212 ms 14. !H. an asterisk (*) is printed for that probe.842 ms 15.20.232.35.44.net (144.sntc01.my-isp-provider.net (144.909 ms 6 sl-gw11-sj-3-0.221.304 ms 14.qxd 2/25/05 9:56 AM Page 73 Using traceroute to Test Connectivity 73 Table 4.232.17) 15.246) 13.120 ms 2 dist3-vlan50.133 ms 15. Communication administratively prohibited. A router Access Control List (ACL) or firewall is in the way.179 ms 4 bb2-p11-0.124.dsl.124. network.net (64.

200.487 ms [root@bigboy tmp]# traceroute Internet Slowness False Alarm The following traceroute gives the impression that a Web site at 80.110) 14.161.133) 16.443 ms 16.221.net (144.6 80.151.246) 16.306 ms 26.200.423 ms 8 sl-bb23-ana-6-0.3.663 ms 5 bb1-p14-0.232.net (67. or packet loss.482 ms 26.124.49) 14.227 Trace complete.158 traceroute to 144.net (63.887 ms 13.my-isp-provider.sntc01.net (144.255.118.200.185 ms 15.168.203.227 1 2 3 4 5 6 7 8 9 10 11 1 43 15 26 38 239 254 24 91 17 30 ms ms ms ms ms ms ms ms ms ms ms 2 15 16 13 12 255 252 20 89 20 16 ms ms ms ms ms ms ms ms ms ms ms 1 44 8 16 14 253 252 20 60 20 23 ms ms ms ms ms ms ms ms ms ms ms 66.44.40.745 ms 4 bb2-p11-0.31.193 64.018 ms 12.20.40.dsl.854 ms 13.net (63.17) 12.another-isp-provider.qxd 2/25/05 9:56 AM Page 74 74 Simple Network Troubleshooting Chapter 4 [root@bigboy tmp]# traceroute -I 144.120.net (144. If any of those conditions existed.05Harrison_ch04.897 ms 15.857 ms 15.snfc21.174.65 64.40.111 ms 2 dist3-vlan50.14 64.12 80.47.net (64.189 ms 7 sl-bb25-sj-6-1. 30 hops max.118.035 ms 13.232.sntc01.203.sntc01. all points past the problematic link would show high latency.my-isp-provider.232.snfc21.200.my-isp-provider.149.618 ms 15.my-isp-provider.161.34 192. Many Internet routing devices give very low priority to traffic related to traceroute in favor of revenue-generating traffic.232.53) 15.164 ms 6 sl-gw11-sj-3-0.158) 27.net (64. .another-isp-provider.96.146 ms 3 bb1-g1-0.064 ms 13.200. latency.67) 13.279 ms 15.20.174.21.97 172.20.229 64.my-isp-provider.227 might be slow because there is congestion along the way at hops 6 and 7 where the response time is over 200ms: C:\>tracert 80.260 ms 15. but not an indication of congestion.20.40.134.250.110 192.35.232.253 192.150. C:\> This indicates only that the devices on hops 6 and 7 were slow to respond with ICMP TTL exceeded messages.124.150.785 ms 17.408 ms 14.228.158 (144. 38 byte packets 1 adsl-67-120-221-110.35.another-isp-provider.118.158).

40.40. Remember.16.1. C:\>tracert 80. latency showed up at hop 7 with the return path being very different. when a user at 64.106.32.25.114 Request timed out. Remember. Possible causes of this problem include the following: ☞ The server has a bad default gateway.65. However.97.100.100. A high traceroute time equates to the round-trip time for both the initial traceroute query to each hop and the response of each hop. but with the server. ☞ The server is running some type of firewall software that blocks traceroute.1 80.25.40. The latency appeared to be caused at hop 10 on the first trace not because that hop was slow. but because that was the first hop at which the return packet traveled back to the source via the congested route. This is because the packet’s return path from the target is sometimes not the same as the path taken to get there. a problem seemed to appear at the 10th hop with OtherNet. The problem is not with the router.18 Tracing route to 80.96.175.200. Always Get a Bidirectional traceroute It is always best to get a traceroute from the source IP to the target IP and also from the target IP to the source IP.106.168.1 65. There was once a routing issue between telecommunications carriers FastNet and SlowNet.qxd 2/25/05 9:56 AM Page 75 Using traceroute to Test Connectivity 75 traceroute Dies at the Router Just Before the Server In this case the last device to respond to the traceroute just happens to be the router that acts as the default gateway of the server.14. you will only receive traceroute responses from functioning devices. traceroute gives the packet round-trip time: .175. ☞ The server is shut down or disconnected from the network. using disguised IP addresses and provider names. the real traffic congestion was occurring where FastNet handed off traffic to SlowNet in the second trace.200 did a traceroute to 40.18 over a maximum of 30 hops 1 33 2 33 3 33 4 47 5 29 6 * 7 ^C C:\> ms ms ms ms ms 49 49 32 32 28 * ms ms ms ms ms 28 28 32 31 32 * ms ms ms ms ms 192. or it has an incorrectly configured NIC.252 80. When a user at IP address 40.05Harrison_ch04.19 81.40.32 did a traceroute to 64. In this case. Here is an example of one such case.16.68.25.

fastnet.snjpca1-br2.snjpca1-br1.24.slownet.175.171.49.inet.another-isp-provider.128.96.253] 3 0 ms 0 ms 0 border-from-40-tesser.26.16.my-isp-provider.webh02-1.24.dnvtco1-br2.32 1 1 ms 1 ms 1 2 1 ms 1 ms 1 3 2 ms 1 ms 1 4 1 ms 1 ms 1 5 2 ms 2 ms 1 [216.25.net [4.another-isp-provider.16.30] ms p4-3.net [64.05Harrison_ch04.25.16.24.slownet.another-isp-provider.61] 13 1031 ms 989 ms 978 [4.another-isp-provider.54] 18 1038 ms 1031 ms 1045 19 1050 ms 1094 ms 1034 20 1050 ms 1094 ms 1034 ms ms ms ms ms [40.126] 5 0 ms 0 ms 0 p3-0.106.144.inet.223.net [4.webhostinc2.65] 7 993 ms 961 ms 999 8 [208.24.11.0.paix-bi1.25.33] 1009 ms 1008 ms 971 ms 9 985 ms 947 ms 983 10 1028 ms 1010 ms 953 11 989 ms 988 ms 985 [4.25.9.26.dnvtco1-br2.96.7.sfj.200 from 40.161] [40.175.another-isp-provider.16.another-isp-provider.net [207.net [4.175.0.another-isp-provider.11.16.26] 17 1030 ms 1020 ms 1045 [4.106.25.11.200] .net sjo-core-01.6.p4-2.5.171.25.net [4.38] 8 30 ms 30 ms 30 p15-0.175.net ms p1-0.200 1 0 ms 0 ms 0 [64.0.another-isp-provider.13] 12 1002 ms 1001 ms 973 [4.150] 10 1252 ms 1212 ms 1202 h0.16.38] 11 1252 ms 1212 ms 1192 [40.net ms p6-0.24.25.slownet.16.25.65] border8.11.another-isp-provider.25.ge0-1-net2.53] 6 0 ms 0 ms 0 p2-1.11] 12 1262 ms 1212 ms 1192 [40.22.snjpca1-br2.another-isp-provider.qxd 2/25/05 9:56 AM Page 76 76 Simple Network Troubleshooting Chapter 4 Trace route to 40.24.96.169] 4 0 ms 0 ms 0 [64.net ms p15-0.another-isp-provider.6.225] 9 30 ms 30 ms 30 p1-0.sfj.3] [40.another-isp-provider.dnvtco1-cr3.52.99] ms [64.19.another-isp-provider.200] 2 0 ms 0 ms 0 [64.37] 16 1045 ms 1037 ms 1050 [4.205.106.24.97] ms [205.96.snjpca1-cr4.net ms sjo-edge-03.106.net [205.net [4.dnvtco1-cr3.128.16.46.236.2] [40.175.net [4.174.77] 6 2 ms 1 ms 1 [216.net ms p3-0.dnvtco1-br1.fastnet.2.net ms gw234.32] Trace route to 64.net [205.162] 13 1102 ms 1091 ms 1092 [40.171.25] 7 0 ms 0 ms 0 p15-0.253] ms [64.net ms p9-0.net ms core1.9.24.16.25.inet.96.29] ms svl-core-03.52.my-isp-provider.226] 15 1027 ms 1025 ms 1023 [4.net ms p0-0.175.130] 14 1031 ms 1017 ms 1017 [4.cointcorp.32 from 64.dnvtco1-br1.

40. transit.153 Pinging 186.net rtr-1. Lost = 0 (0% loss).64.94: 186.40.94] [186.93] [186.17.net [186.93] [186.40..64.9. at which point you get the timeout.40.17.64.. Each bounce causes the TTL to decrease by a count of 1 until the TTL reaches 0.my-isp-provider.93] This problem was solved by resetting the routing process on both routers.9. ms ms ms ms ms ms ms ms ms <10 60 70 60 70 61 70 60 70 ms ms ms ms ms ms ms ms ms 186.net rtr-2.64. .94] [186. traceroute Web Sites Many ISPs will provide their subscribers with the facility to do a traceroute from purpose-built servers called looking glasses.64.net rtr-1.93: G:\>ping 186.9. Approximate round trip times in milli-seconds: Minimum = 0ms. Ping statistics for 186.64.17.net rtr-1.153 gave a TTL timeout message.40.153: Packets: Sent = 4.64.64.net rtr-2.40.94] [186. . in which the packet was proven to be bouncing between routers at 186.40.17.40. The problem was initially triggered by an unstable network link that caused frequent routing recalculations.my-isp-provider.40..94: TTL TTL TTL TTL expired expired expired expired in in in in transit.9.my-isp-provider.my-isp-provider.9.64.my-isp-provider. The constant activity eventually corrupted the routing tables of one of the routers.64. transit.17.40.9.1 rtr-2.my-isp-provider.qxd 2/25/05 9:56 AM Page 77 Using traceroute to Test Connectivity 77 ping and traceroute Troubleshooting Example In this example.64. Average = 0ms G:\>tracert 186. Doing a ..153 Tracing route to lostserver. Received = 4. transit.94 and 186.net [186. The routing loop was confirmed by the traceroute.217.net rtr-1. Maximum = 0ms.my-isp-provider..40.93] [186.my-isp-provider.40.net rtr-2.my-isp-provider.40. Ping TTLs will usually timeout only if there is a routing loop in which the packet bounces between two routers on the way to the target.33.05Harrison_ch04.94] [186.94: 186.64. a ping to 186.64. A simple Web search for the phrase Internet looking glass will provide a long list of alternatives. Trace complete.17.40.64..94: 186.153] over a maximum of 30 hops: 1 <10 ms <10 2 60 ms 70 3 70 ms 71 4 60 ms 70 5 70 ms 70 6 60 ms 70 7 70 ms 70 8 60 ms 70 9 70 ms 70 .153 with 32 bytes of data: Reply Reply Reply Reply from from from from 186.

☞ Do a traceroute from this router to your intended target server. It’s usually good to do the following traceroute ☞ ☞ ☞ ☞ ☞ ☞ Log on to the last visible router.qxd 2/25/05 9:56 AM Page 78 78 Simple Network Troubleshooting Chapter 4 traceroute from a variety of locations can help identify whether the problem is with the ISP of your Web server or the ISP used at home/work to provide you with Internet access. ☞ If it doesn’t work: Test the routing table and/or other status of all the hops between it and your intended target. The last visible hop is the last hop in which the packets return correctly. Do a traceroute back to your source server. (!H or !N messages may be produced. Possible Reasons for a Failed traceroute A traceroute can fail to reach its intended destination for a number of reasons including the following: ☞ packets are being blocked or rejected by a router in the path. (!H or !N messages may be produced.05Harrison_ch04. which provides a list of looking glasses sorted by country. The packets don’t have a proper return path to your server. It’s usually good to check the routing table and/or other status of this next hop device. . ☞ Look at the routing table to determine what the next hop is to your intended traceroute target. You may have a routing loop in which packets bounce between two routers and never get to the intended destination. The router immediately after the last visible one is the one at which the routing changes. The router immediately after the last visible one is usually the culprit. Note If there is nothing blocking your traceroute traffic. ☞ If this works: Routing to the target server is OK.org. A more convenient way of doing this is to use a site like traceroute. It could be disconnected or turned off. The target server doesn’t exist on the network.) You may have a typographical error in the IP address of the target server. ☞ Log on to this next hop router. the last visible router of an incomplete trace is either the last good router on the path or the last router that has a valid return path to the server issuing the traceroute. The traceroute will probably fail at the bad router on the return path.) The network on which you expect the target host to reside doesn’t exist in the routing table of one of the routers in the path.

Restart statistics Q . and average round-trip times in milliseconds for the probe packets between each hop along the way to the final destination. “Installing RPM Software.168. You type mtr followed by the target IP address to get output similar to the following: [root@bigboy tmp]# mtr 192.168.18.35 0% 16 16 24 23 29 42 5.” There is even a free Windows version called WinMTR.52] Bigboy Fri Feb 20 17:19:17 2004 Keys: D . 192. MTR is automatically installed as part of Fedora Linux.25. VIEWING PACKET FLOWS WITH TCPDUMP The tcpdump command is one of the most popular packages for viewing the flow of packets through your Linux box’s NIC card. 192. but also to determine when they appear to be slow.168.Quit Packets Pings Hostname %Loss Rcv Snt Last Best Avg Worst 1.Display mode R .05Harrison_ch04.168. 192.26 Matt’s traceroute [v0.3. The constant refreshing of the screen also enables you to instantaneously spot transient changes in quality fairly easily. The installation of RPMs is covered in Chapter 6.1.15 0% 17 17 23 14 18 25 4.25.qxd 2/25/05 2:35 PM Page 79 Viewing Packet Flows with tcpdump 79 USING MTR TO DETECT NETWORK CONGESTION Matt’s Traceroute is an application you can use to do a repeated traceroute in real time. you can download the RPM software installation package from many of the Fedora download sites. worst. . The constant updates enable you not only to visually determine which hops are slow.254 0% 17 17 12 11 18 41 3. It is installed by default on Red Hat/Fedora Linux and has very simple syntax. If MTR isn’t installed on your system. it dynamically shows the round-trip time to reach each hop along the traceroute path. making it much more convenient than a regular traceroute.1 0% 17 17 32 10 15 32 2. The advantage of this is that you can let MTR run for an extended period of time. 192. It is a good tool to use whenever you suspect there is some intermittent network congestion.168. acting as a constant monitor of communication path quality. 192.2.26 0% 16 16 23 21 26 37 ^C [root@bigboy tmp]# One of the nice features of MTR is that it gives you the best.168. especially if you are doing simpler types of troubleshooting.

Like most Linux commands. Some useful expressions are listed in Table 4. Table 4. These act as filters to limit the volume of data presented on the screen. tcpdump uses command-line switches to modify the output.2. Table 4. Some of the more useful command-line switches are listed in Table 4. If this is not specified.3.qxd 2/25/05 9:56 AM Page 80 80 Simple Network Troubleshooting Chapter 4 One of the most common uses of tcpdump is to determine whether you are getting basic two-way communication.3 Tcpdump Useful tcpdump Expressions Description Command Expression host host-address icmp tcp port port-number udp port port-number View packets from the IP address host-address. View icmp packets. View UDP packets with either a source or destination UDP port of port-number. Specify the size the dump file must reach before a new one with a numeric extension is created. .2 Tcpdump Possible tcpdump Switches Command -c -t -i -w -C Switch Description Stop after viewing count packets. Dump the output to a specially formatted tcpdump dump file. common culprits are firewalls. You can also use keywords such as and or or between expressions to further fine-tune your selection criteria.05Harrison_ch04. Don’t print a timestamp at the beginning of each line. tcpdump uses the lowest numbered interface that is UP. You can also add expressions after all the command-line switches. Lack of communication could be due to the following: ☞ Bad routing ☞ Faulty cables. routers with access control lists and even your Linux box running iptables Analyzing tcpdump in much greater detail is beyond the scope of this section. View TCP packets with either a source or destination TCP port of port-number. Listen on interface. interfaces of devices in the packet flow ☞ The server not listening on the port because the software isn’t installed or started ☞ A network device in the packet path blocking traffic.

32938: R smallfry.sackOK.wscale 0> (DF) [tos 0x10] .927091 smallfry > bigboy: icmp: echo request (DF) 21:48:58. > smallfry: icmp: echo reply 21:48:58.05Harrison_ch04.962244 bigboy > smallfry: icmp: echo reply 21:48:58.1.168.1.927510 bigboy > smallfry: icmp: echo reply 21:48:58.102 and tcp port 22 tcpdump: listening on wlan0 smallfry.timestamp 75227931 bigboy.qxd 2/25/05 9:56 AM Page 81 Viewing Packet Flows with tcpdump 81 The following is an example of tcpdump being used to view ICMP packets going through interface wlan0: [root@bigboy tmp]# tcpdump -i wlan0 icmp ping tcpdump: listening on wlan0 21:48:58.32938 > bigboy.928365 bigboy.32938 > bigboy.my-web-site.968556 bigboy > smallfry: icmp: echo reply 9 packets received by filter 0 packets dropped by kernel [root@bigboy tmp]# In this example: ☞ The first column of data is a packet timestamp. ☞ The second column of data shows the packet source and then the destination IP address or server name of the packet.168.ssh: S 1460.sackOK.nop. ☞ The third column shows the packet type.944034 bigboy > smallfry: icmp: echo reply 21:48:58. ☞ Two-way communication is occurring as each echo gets an echo reply.ssh: S 1460.928257 smallfry > bigboy.943926 smallfry > bigboy.nop.102 on TCP port 22 with no timestamps in the output (-t switch): [root@bigboy tmp]# tcpdump -i wlan0 -t host 192.org: icmp: echo request (DF) 21:48:58. The following example shows tcpdump being used to view packets on interface wlan0 to/from host 192.32938 > bigboy.my-web-site.sackOK.32938: R 0x10] smallfry.timestamp 75227931 7 packets received by filter 0 packets dropped by kernel [root@bigboy tmp]# 2013297020:2013297020(0) win 5840 <mss 0.ssh: S 1460.wscale 0> (DF) [tos 0x10] 0:0(0) ack 1 win 0 (DF) [tos 0x10] 2013297020:2013297020(0) win 5840 <mss 0.wscale 0> (DF) [tos 0x10] 0:0(0) ack 2013297021 win 0 (DF) [tos 2013297020:2013297020(0) win 5840 <mss 0.org: icmp: echo request (DF) 21:48:58.nop.963966 bigboy > smallfry: icmp: echo reply 21:48:58.timestamp 75227931 bigboy.ssh > smallfry.ssh > smallfry.

ssh > 192-168-1-242. Analyzing tcpdump files By using the -w filename option you can send the entire Ethernet frame.org.ssh: .019118 IP bigboy-100.my-web-site. and the capability of automatically highlighting certain error conditions such as data retransmissions: tcpdump -i eth1 -w /tmp/packets.1753: P 165:401(236) ack 0 win 6432 .myweb-site.dump tcp port 22 Covering Ethereal is beyond the scope of this book.myweb-site.818478 IP bigboy-100. This can slow down tcpdump so much that it appears not to be working at all. The -n switch stops DNS name lookups and makes tcpdump work more reliably.1753: P 1:165(164) ack 0 win 6432 02:24:35. which is available in both Windows and Linux. capture size 96 bytes 02:24:34. and a Windows version is also available. ☞ The client named bigboy is using port 32938 to communicate with the server named smallfry on the TCP SSH port 22. Common Problems with tcpdump By default tcpdump will attempt to determine the DNS names of all the IP addresses it sees while logging data.ssh > 192-168-1-242.1753 > bigboy-100.myweb-site.org.org. Without the -n switch: [root@bigboy tmp]# tcpdump -i eth1 tcp port 22 tcpdump: verbose output suppressed. use -v or -vv for full protocol decode listening on eth1.myweb-site.my-web-site. The following are examples of how the -n switch affects the output.1753 > bigboy-100.org. not just brief IP information that normally goes to the screen.org.818398 IP 192-168-1-242.05Harrison_ch04. but that shouldn’t discourage you from using it. link-type EN10MB (Ethernet).org.org.my-web-site.019042 IP 192-168-1-242.qxd 2/25/05 9:56 AM Page 82 82 Simple Network Troubleshooting Chapter 4 In this example: ☞ The first column of data shows the packet source and then the destination IP address or server name of the packet.ssh: .org. The application is part of the Fedora RPM suite. ☞ Two-way communication is occurring. colorization of packet records based on criteria deemed interesting. to a file. with customized filters. ☞ The second column shows the TCP flags within the packet. ack 165 win 65307 02:24:35. ack 318574223 win 65471 02:24:34.my-web-site. This can then be analyzed by graphical analysis tools such as Ethereal.

link-type EN10MB (Ethernet). though the dump file format is identical to tcpdump. which is also known as a ring buffer.1.168.068606 IP 192.org.242.org.068511 IP 192.242.ssh: .ssh > 192.100.1.ssh > 192.100. ack 165 win 64999 02:25:53.ssh > 192-168-1-242.242.ssh: P 0:20(20) ack 401 win 65071 02:24:35.ssh > 192.100.168.1.168.myweb-site.242.100.my-web-site.176299 IP 192-168-1-242.1.myweb-site.5 show some popular command switches and expressions that can be used with tethereal.1753 > 192.168. The tethereal command has the capability of dumping data to a file like tcpdump and creating new files with new filename extensions when a size limit has been reached. ack 318576011 win 65163 02:25:53.168.1. The command-line options and screen output mimic that of tcpdump in many ways.176337 IP bigboy-100.168. .ssh: . capture size 96 bytes 02:25:53.my-web-site.168.4 and 4.1.qxd 2/25/05 9:56 AM Page 83 Viewing Packet Flows with tethereal 83 02:24:35.1. It can additionally limit the total number of files created before overwriting the first one in the queue.1753 > 192.05Harrison_ch04.168.1753 > bigboy-100.242.1753 > 192.1.100.100. but tethereal has a number of advantages.1753: P 401:629(228) ack 20 win 6432 6 packets captured 7 packets received by filter 0 packets dropped by kernel [root@bigboy tmp]# With the -n switch: [root@bigboy tmp]# tcpdump -i eth1 -n tcp port 22 tcpdump: verbose output suppressed.1.1753: P 353:541(188) ack 20 win 6432 6 packets captured 7 packets received by filter 0 packets dropped by kernel [root@bigboy tmp]# VIEWING PACKET FLOWS WITH TETHEREAL The tethereal program is a text version of the graphical Ethereal product that is part of the Fedora Linux RPM suite.242.408556 IP 192. Tables 4.1. The tethereal screen output is also more intuitive to read.269205 IP 192.1.168.org.org.168. use -v or -vv for full protocol decode listening on eth1.168.168.408589 IP 192.1753: P 165:353(188) ack 0 win 6432 02:25:53.1753: P 1:165(164) ack 0 win 6432 02:25:53.269152 IP 192.ssh: P 0:20(20) ack 353 win 64811 02:25:53.1.

100. Specify the size the dump file must reach before a new one with a numeric extension is created.5 tethereal Command Expression host host-address Icmp tcp port port-number udp port port-number In the next example we’re trying to observe an HTTP (TCP port 80) packet flow between server smallfry at address 192.1.168.168.168.100 -> 192.000000 192.168.100 -> 192.1.100 TCP 1442 Seq=3325831829 Ack=3291904937 Win=5840 Len=0 2.168.100 TCP 1442 Seq=3325831829 Ack=3291904937 Win=5840 Len=0 2.102 TCP http Seq=3291904937 Ack=3325831830 Win=46 Len=0 2.102 -> 192.1.168.100 Capturing on eth0 0.168.100 TCP 1442 Seq=3325831830 Ack=3291904938 Win=5840 Len=0 [root@smallfry tmp]# and host > http [SYN] > 1442 [SYN.1. tethereal Command Switch -c -i -w -C -b Table 4. tethereal will use the lowest numbered interface that is UP.1.102 -> 192.100 -> 192. If this is not specified.1.603356 192.102 and bigboy at IP address 192.102 -> 192. The tethereal output groups the IP addresses and TCP ports together and then provides the TCP flags.603211 192.168.168. View icmp packets.168.1.100 TCP 1442 Seq=3325831828 Ack=0 Win=5840 Len=0 0.1.102 TCP http Seq=3291904937 Ack=3325831830 Win=46 Len=0 2. followed by the sequence numbering.1.1.168. Useful tethereal Expressions Description View packets from the IP address host-address.05Harrison_ch04.168.000223 192. ACK] > http [ACK] . View UDP packets with either a source or destination UDP port of port-number. Listen on interface. but the formatting lines up in neat columns on your screen.168. It may not be apparent on this page.1. Also notice how the command line mimics that of tcpdump: [root@smallfry tmp]# tethereal -i eth0 tcp port 80 192.1. Determine the size of the ring buffer when the -C switch is selected.168.168.602804 192.168.4 Possible tethereal Switches Description Stop after viewing count packets. making analysis much easier.102 -> 192.1.000157 192. ACK] > 1442 [ACK] > 1442 [FIN.1.603398 192.qxd 2/25/05 9:56 AM Page 84 84 Simple Network Troubleshooting Chapter 4 Table 4.1.102 TCP http Seq=3291904936 Ack=3325831829 Win=5792 Len=0 0.1.168. Dump the output to a specially formatted tethereal dump file. ACK] > http [ACK] > http [FIN. View TCP packets with packets either a source or destination TCP port of port-number.1.

. Using nslookup to Test DNS The nslookup command can be used to get the associated IP address for your domain and vice versa. but that shouldn’t discourage you from using it. The application is part of the Fedora RPM suite and a Windows version is also available.com: the IP address 216. .193. .193. as explained in Chapter 18.151.com Address: 216.com .151.linuxhomenetworking..151...linuxhomenetworking.92 ..arpa name = extra193-92.193.conf file. Name: www.. you just need to type the command. which is then displayed on the screen.linuxhomenetworking.151.qxd 2/25/05 9:56 AM Page 85 Basic DNS Troubleshooting 85 Using graphical Ethereal to analyze tethereal dump files is beyond the scope of this book. BASIC DNS TROUBLESHOOTING Sometimes the source of problems can be due to misconfigured DNS rather than poor network connectivity. Using nslookup to Check Your Web Site Name Here we see nslookup returning www.193.in-addr. Failures can be caused by your server not having the correct value set in the /etc/resolv.. The command actually queries your DNS server for a response. The nslookup command is very easy to use.151.my-isp-provider.05Harrison_ch04. As mentioned before.92 for the site [root@bigboy tmp]# nslookup www.com: [root@bigboy tmp]# nslookup 216.193. followed by the IP address or Web site name you want to query.. poor connectivity to your DNS server.. or an incorrect configuration on the DNS server.92 [root@bigboy tmp]# Using nslookup to Check Your IP Address The nslookup command can operate in the opposite way in which a query against the address 216.net.216. DNS is the system that helps map an IP address to your Web site’s domain name and your site may suddenly become unavailable if the mapping is incorrect.92 returns the Web site named www.linuxhomenetworking. Non-authoritative answer: 92.

linuxhomenetworking. syntax is identical to that of nslookup and the resulting output is very similar: [root@bigboy tmp]# host 216.92 92.96.34 [root@zippy root]# .151..net internet address = 216.87.87.151. .1 [root@bigboy tmp]# Using nslookup to Query a Specific DNS Server Sometimes you might want to test the DNS mapping against a specific DNS server.3 Address: 68.qxd 2/25/05 9:56 AM Page 86 86 Simple Network Troubleshooting Chapter 4 Authoritative answers can be found from: 193.linuxhomenetworking.in-addr.com has address 65.05Harrison_ch04. The Linux host command can be used as a good replacement.3 Using domain server: Name: 68.193..92 [root@bigboy tmp]# [root@zippy root]# host www.151.87.96. 193.115..216.com www.com has address 216.my-isp-provider.3 .arpa nameserver = dns2.in-addr.193.3#53 Aliases: www. [root@bigboy tmp]# [root@bigboy tmp]# host www.my-isp-provider.arpa domain name pointer extra193-92.3#53 Name: www.com Address: 216.3 Address: 68.net.net.linuxhomenetworking.193. but can still be used with Windows.96. Fortunately.87.linuxhomenetworking.87.linuxhomenetworking.216.linuxhomenetworking.216. Server: 68.arpa nameserver = dns1.151..71. this can be achieved by adding the DNS server’s IP address immediately after the IP address of the Web site name you intend to query: [root@bigboy tmp]# nslookup www.my-isp-provider.96.96.com 68.com 68.96.151.193.92 [root@bigboy tmp]# Note The nslookup command will probably be removed from future releases of Linux.151.87. Using the host Command to Test DNS More recent versions of Linux have started to use the host command for basic DNS testing.192. dns1.in-addr.my-ispprovider.net.151.

This option can be used to send them from 60 seconds apart with a value of 5.00 ( www. 3..6.” to 0. .00 Usage: nmap [Scan Type(s)] [Options] <host or net list> Some Common Scan Types (‘*’ options require root privileges) * -sS TCP SYN stealth port scan (default if privileged (root)) -sT TCP connect() port scan (default for unprivileged users) * -sU UDP port scan -sP ping scan (Find any reachable machines) .3 seconds with a value of 0 in “paranoid mode. It is a favorite tool of malicious surfers and therefore should be used to test external as well as internal servers under your control.com (192. Table 4. [root@bigboy tmp]# Some of the more common nmap options are listed in Table 4.6 Argument -P0 -T Commonly Used nmap Options Description Attempts to ping a host before scanning it. but you should also refer to the nmap man pages for full descriptions of them all. Defines the timing between the packets set during a port scan.1. If the server is being protected from ping queries.153): (The 4981 ports scanned but not shown below are in state: closed) Port State Service . you can get a list of available nmap options by just entering the command without arguments at the command prompt: [root@bigboy tmp]# nmap Nmap V. you can use this option to force it to scan anyway..153 Starting nmap V.. Defines a variety of scan methods that use either packets that comply with the TCP/IP standard or are in violation of it.” Tries to detect the operating system of the remote server based on known responses to various types of packets. such as servers running unauthorized network applications.168. but it can be used in a corporate environment to detect vulnerabilities in your network.1. “insane mode. Some firewalls can detect the arrival of too many nonstandard packets within a predetermined time frame.insecure.05Harrison_ch04.qxd 2/25/05 9:56 AM Page 87 Using nmap 87 USING NMAP You can use nmap to determine all the TCP/IP ports on which a remote server is listening. 3.org/nmap/ ) Interesting ports on whoknows. Lists the TCP/IP port range to scan. Whenever you are in doubt.my-site-int.168. It isn’t usually an important tool in the home environment. -O -p -s Here is an example of trying to do a scan using valid TCP connections (-sT) in the extremely slow insane mode (-T 5) from ports 1 to 5000: [root@bigboy tmp]# nmap -sT -T 5 -p 1-5000 192..

but of the type that you can’t handle.. but though you may have the IP address. One of the better ones to use is the whois command. This could be either malicious or simply someone rapidly downloading all the pages of your Web site with the Linux wget command.05Harrison_ch04.com +1.com domain: [root@bigboy tmp]# whois yahoo. . Use it with an IP address or DNS domain as its sole argument and it will provide you with all the administrative information you need to start your hunt.. Symptoms include a large number of established connections when viewed with the netstat command or an excessive number of entries in your firewall or Web server logs. ISPs are usually very sensitive to complaints about SPAM.4083493300 . Sometimes DNS lookups aren’t enough to determine who owns an offending IP address. Here is an example for the yahoo. a traceroute won’t provide any contact information for the ISP.. You need another tool. Administrative Contact: Domain Administrator (NIC-1382062) Yahoo! Inc. but you should go the extra mile and purchase a text specifically on Linux security to help protect you against attempts at malicious security breaches.com .1 IP address (1 host up) scanned in 8 seconds [root@bigboy tmp]# Full coverage of the possibilities on nmap as a security scanning tool are beyond the scope of this book.qxd 2/25/05 9:56 AM Page 88 88 Simple Network Troubleshooting Chapter 4 21/tcp 25/tcp 139/tcp 199/tcp 2105/tcp 2301/tcp 3300/tcp open open open open open open open ftp smtp netbios-ssn smux eklogin compaqdiag unknown Nmap run completed -. 701 First Avenue Sunnyvale CA 94089 US domainadmin@yahoo-inc. such as e-mail SPAM.. DETERMINING THE SOURCE OF AN ATTACK Sometimes you realize that your system is under a denial-of-service attack. Sometimes the attack isn’t in the form of a constant bombardment that your server can’t handle.

org logged into bigboy as user root.05Harrison_ch04. Here we see a user logged as root from server reggae.+1. which lists the last users who logged into the system.my-web-site. .4.09:26 (1+08:18) reboot system boot 2. [root@bigboy tmp]# WHO HAS USED MY SYSTEM? It is always important to know who has logged into your Linux box.org Mon Jun 16 07:24 . The who Command The who command is used to see who is currently logged in to your computer. The last Command The most common command to determine who has logged into your system is last.qxd 2/25/05 9:56 AM Page 89 Who Has Used My System? 89 Fax.my-web-site.down (03:07) root pts/0 reggae.4083493301 . Here are some examples: [root@bigboy tmp]# last -100 root pts/0 reggae.my-web-site.my-web-site..org Wed Jun 18 01:07 . This makes it easier to track down individuals rather than groups of users.my-web-site.org Tue Jun 17 21:57 . They can then get root privileges by using sudo...conf file.. but mostly to figure out who made the mistake that crashed the system or blew up Apache with a typographical error in the httpd.org: [root@bigboy tmp]# who root pts/0 Jun 19 09:26 (reggae.00:35 (17:10) wtmp begins Sun Jun 15 16:29:18 2003 [root@bigboy tmp]# In this example someone from reggae. This isn’t just to help track the activities of malicious users. I generally prefer not to give out the root password and let all the systems administrators log in with their own individual logins.my-web-site.org) [root@bigboy tmp]# .org Thu Jun 19 09:26 still logged in root pts/0 reggae.my-web-site.18-14 Wed Jun 18 01:07 (1+08:21) root pts/0 reggae.

The next chapter builds on this new knowledge and expands your troubleshooting skills to include the reading of Linux error log files to assist in the diagnosis of unexpected Linux application behavior. These tips should help provide you with a definitive answer by enabling you to isolate the source of most network problems and helping you make their resolution much faster. especially as network instability is often under the control of network engineers who need evidence pointing to problems in their domain of expertise before they will be convinced to act.05Harrison_ch04. . The worry can be amplified.qxd 2/25/05 9:56 AM Page 90 90 Simple Network Troubleshooting Chapter 4 CONCLUSION One of the greatest sources of frustration for any systems administrator is to try to isolate whether poor server performance is due to a network issue or problems with an application or database.

and Web searches. online documentation. but will usually log them somewhere. Linux applications use the syslog utility to export all their errors and status messages to files located in the /var/log directory. This can be invaluable in correlating the timing and causes of related events on your system. 91 . I’ve included syslog here as a dedicated chapter to both emphasize its importance to your Linux knowledge and prepare you with a valuable skill that will help you troubleshoot all the Linux various applications that will be presented throughout the book.qxd 2/25/05 9:57 AM Page 91 C H A P T E R 5 Troubleshooting Linux with syslog IN THIS CHAPTER ☞ syslog ☞ logrotate ☞ Conclusion There are hundreds of Linux applications on the market. Fortunately. It is also important to know that applications frequently don’t display errors on the screen. in most cases. each with their own configuration files and help pages. but it also makes Linux system administration daunting. syslog and the logrotate utility that cleans up log files are both relatively easy to configure but they frequently don’t get their fair share of coverage in most texts.06Harrison_ch05. Knowing the precise message that accompanies an error can be vital in researching malfunctions in product manuals. This variety makes Linux vibrant.

Table 5.none.qxd 2/25/05 2:36 PM Page 92 92 Troubleshooting Linux with syslog Chapter 5 SYSLOG syslog is a utility for tracking and logging all manner of system messages from the merely informational to the extremely critical. applications such as mail and cron generate messages with easily identifiable facilities named mail and cron.1 syslog Facilities Severity Level 0 1 2 3 4 5 6 7 Keyword emergencies alerts critical errors warnings notifications informational debugging Description System unusable Immediate action required Critical condition Error conditions Warning conditions Normal but significant conditions Informational messages Debugging messages The /etc/syslog. all messages of severity “info” and above are logged.1. or authentication facilities/subsystems.authpriv. You can configure syslog’s /etc/syslog.conf configuration file to place messages of differing severity and facilities in different files. cron. Each system message sent to the syslog server has two descriptive labels associated with it that makes the message easier to handle: ☞ The first describes the function (facility) of the application that generated it.cron. Here is a sample: . This file consists of two columns: The first lists the facilities and severity of messages to expect. This procedure will be covered next. but none from the mail.none.none /var/log/messages In this case.info. You can make this logging even more sensitive by replacing the line above with one . For example. Red Hat/Fedora’s /etc/ syslog. There are eight in all and they are listed in Table 5.conf file is configured to put most of the messages in the file /var/log/messages.06Harrison_ch05.conf configuration file.mail. ☞ The second describes the degree of severity of the message. By default. and the second lists the files to which they should be logged.conf File The files to which syslog writes each type of message received is set in the /etc/syslog.

The use of spaces in the file would cause unpredictable results.conf will not take effect until you restart syslog. Activating Changes to the syslog Configuration File Changes to /etc/syslog.conf file to be safe. you can use this command: [root@bigboy tmp]# tail -f /var/log/messages .conf file. Here are some common examples: Files /var/log/maillog /var/log/httpd/access_log : Mail : Apache web server page access logs Directories /var/log /var/log/samba /var/log/mrtg /var/log/httpd : Samba messages : MRTG messages : Apache webserver messages Note In some older versions of Linux the /etc/syslog. Issue this command to do so: [root@bigboy tmp]# service syslog restart How to View New Log Entries as They Happen If you want to get new log entries to scroll on the screen as they occur.qxd 2/25/05 9:57 AM Page 93 syslog 93 that captures all messages from debug severity and above in the messages file.06Harrison_ch05. This may be more suitable for troubleshooting: *.debug /var/log/ /var/log/messages Certain applications will additionally log to their own application specific log files and directories independent of the syslog. Check the formatting of your /etc/syslog.conf file was very sensitive to spaces and would recognize only tabs.

Configuring the Linux syslog Server By default. syslog checks its /etc/syslog. syslog will not listen for remote messages unless the SYSLOGD_OPTIONS variable in this file has an -r included in it: # Options to syslogd # -m 0 disables ‘MARK’ messages. It also makes covering up mistakes or malicious activities harder because the purposeful deletion of log files on a server cannot simultaneously occur on your logging server. As we saw previously. once for klogd to decode. Another good command to use apart from tail is grep. and # once for processing with ‘ksymoops’ # -x disables all klogd processing of oops messages entirely # See klogd(8) for more details KLOGD_OPTIONS=”-2” .conf file to determine the expected names and locations of the log files it should create. Here’s how to configure your Linux server to start listening for these messages. especially if you restrict the user access to the logging server. # -r enables logging from remote machines # -x disables DNS lookups on messages received with -r # See syslogd(8) for more details SYSLOGD_OPTIONS=”-m 0 -r” # Options to klogd # -2 prints all kernel oops messages twice.qxd 2/25/05 9:57 AM Page 94 94 Troubleshooting Linux with syslog Chapter 5 Similar commands can be applied to all log files. Here is an example: [root@bigboy tmp]# more /var/log/messages Logging syslog Messages to a Remote Linux Server Logging your system messages to a remote server is a good security practice. grep will help you search for all occurrences of a string in a log file.06Harrison_ch05. Here is an example: [root@bigboy tmp]# grep string /var/log/messages | more You can also just use the plain old more command to see one screen at a time of the entire log file without filtering with grep. syslog doesn’t expect to receive messages from remote clients. It also checks the file /etc/sysconfig/syslog to determine the various modes in which it should operate. This is probably one of the best troubleshooting tools available in Linux. it becomes easier to correlate events across your company. With all servers logging to a central syslog server. you can pipe it through the more command so that you only get one screen at a time.

debug *.0:514 0.debug @loghost /var/log/messages You have now configured all debug messages and higher to be logged to both server bigboy (“loghost”) and the local file /var/log/messages. The server will now start to listen on UDP port 514. which you can verify using either one of the following netstat command variations: [root@bigboy tmp]# netstat -a | grep syslog udp 0 0 *:syslog *:* [root@bigboy tmp]# netstat -an | grep 514 udp 0 0 0.0.0:* [root@bigboy tmp]# Configuring the Linux Client The syslog server is now expecting to receive syslog messages. Remember to restart syslog to get the remote logging started.06Harrison_ch05.conf file to make the syslog messages get sent to your new loghost nickname: *.0. such as restarting the lpd printer daemon and making sure the remote server sees the messages. This is done by editing the /etc/hosts file on the Linux client named smallfry.168. Determine the IP address and fully qualified hostname of your remote logging host. Linux Client [root@smallfry tmp]# service lpd restart Stopping lpd: [ OK ] Starting lpd: [ OK ] [root@smallfry tmp]# . 2.0. You can now test to make sure that the syslog server is receiving the messages with a simple test. Here are the steps: 1.1.0.qxd 2/25/05 9:57 AM Page 95 syslog 95 You have to restart syslog on the server for the changes to take effect.org bigboy loghost Now your /etc/hosts file has a nickname of “loghost” for server bigboy. 3. Add an entry in the /etc/hosts file in the format: IP-address fully-qualified-domain-name hostname “loghost” Example: 192.my-web-site. The next thing you need to do is edit your /etc/syslog.100 bigboy. You have to configure your remote Linux client to send messages to it.

.. and load balancers—each logging with a different facility—can each have their own log files for easy troubleshooting.06Harrison_ch05. If you are logging to a remote syslog server via a firewall. Appendix IV has examples of how to configure syslog to do this with Cisco devices using separate log files for the routers.conf file is logrotate’s general configuration file in which you can specify the frequency with which the files are reused: ☞ You can specify either a weekly or daily rotation parameter. [root@bigboy tmp]# syslog Configuration and Cisco Network Devices reserves facilities local0 through local7 for log messages received from remote servers and network devices. In the case below.. ☞ The create parameter creates a new log file after each rotation. while allowing 7 copies. . the weekly option is commented out with a #. In the case below. switches.. . LOGROTATE The Linux utility logrotate renames and reuses system error log files on a periodic basis so that they don’t occupy excessive disk space. firewalls. you have to allow traffic on this port to pass through the security device. switches. . ☞ The rotate parameter specifies the number of copies of log files logrotate will maintain... and LocalDirectors.conf File The /etc/logrotate. syslog messages usually have UDP port 514 for both their source and destination UDP ports. allowing daily updates. the 4 copy option is commented out with a #.. Apr 11 22:09:35 smallfry lpd: lpd shutdown succeeded Apr 11 22:09:39 smallfry lpd: lpd startup succeeded . The /etc/logrotate.qxd 2/25/05 9:57 AM Page 96 96 Troubleshooting Linux with syslog Chapter 5 Linux Server [root@bigboy tmp]# tail /var/log/messages . syslog syslog and Firewalls syslog listens by default on UDP port 514. PIX firewalls. Routers.. CSS load balancers.

our sample configuration file will create daily archives of all the logfiles and store them for seven days. The files will have the following names.1 logfile.3 logfile. Here are some sample files that define the specific files to be rotated for each application.pid 2> /dev/null` 2> /dev/null || true endscript } .d Directory Most Linux applications that use syslog put an additional configuration file in this directory to specify the names of the log files to be rotated. It is a good practice to verify that all new applications that you want to use the syslog log have configuration files in this directory.2 logfile.qxd 2/25/05 9:57 AM Page 97 logrotate 97 Therefore.5 logfile. The /etc/logrotate.6 Sample Contents of /etc/logrotate. with logfile the current active version: logfile logfile.d/syslog File (for General System Logging) /var/log/messages /var/log/secure /var/log/maillog /var/log/spooler /var/log/boot.log /var/log/cron { sharedscripts postrotate /bin/kill -HUP `cat /var/run/syslogd.06Harrison_ch05.4 logfile.0 logfile.conf # rotate log files weekly #weekly # rotate log files daily daily # keep 4 weeks worth of backlogs #rotate 4 # keep 7 days worth of backlogs rotate 7 # create new (empty) log files after rotating old ones create The /etc/logrotate.

pid 2> /dev/null` 2> /dev/null || true endscript } Activating logrotate The logrotate settings in the last section will not take effect until you issue the following command: [root@bigboy tmp]# logrotate -f If you want logrotate to reload only a specific configuration file.d/syslog Compressing Your Log Files On busy Web sites the size of your log files can become quite large.log { notifempty missingok sharedscripts copytruncate postrotate /bin/kill -HUP `cat /var/lock/samba/*.d/samba File (for Samba) /var/log/samba/*.d/apache File (for Apache) /var/log/httpd/access_log /var/log/httpd/agent_log /var/log/httpd/error_log /var/log/httpd/referer_log { missingok sharedscripts postrotate /bin/kill -HUP `cat /var/run/httpd.qxd 2/25/05 9:57 AM Page 98 98 Troubleshooting Linux with syslog Chapter 5 The /etc/logrotate. issue the logrotate command with just that filename as the argument: [root@bigboy tmp]# logrotate -f /etc/logrotate. .06Harrison_ch05.conf file and adding the compress option. and not all of them.pid 2>/dev/null` 2> /dev/null || true endscript } The /etc/logrotate. Compression can be activated by editing the logrotate.

.info.0.gz .1.7.0.info.1.6.5. This syslog chapter should make you more confident to learn more about these applications via experimentation because you’ll at least know where to look at the first sign of trouble.2. [root@bigboy tmp]# CONCLUSION In the next chapter we cover the installation of Linux applications.. dsn=4.conf # # Activate log compression compress The log files will then start to become archived with the gzip utility. .info>.34]. [222. and finally. each file having a .3.gz /var/log/messages. the Apache Web server. pri=6388168. whose configuration files are probably the most frequently edited system documents of all and which correspondingly suffer from the most mistakes. Linux mail.4. mailer=esmtp.clematis4spiders. [root@bigboy tmp]# ls /var/log/messages* /var/log/messages /var/log/messages. Nov 15 04:08:02 bigboy httpd: httpd shutdown succeeded Nov 15 04:08:04 bigboy httpd: httpd startup succeeded Nov 15 04:08:05 bigboy sendmail[6003]: iACFMLHZ023165: to=<tvaughan@clematis4spiders. which can be configured to ignore and then log all undesirable packets.134.qxd 2/25/05 9:57 AM Page 99 Conclusion 99 # # File: /etc/logrotate.gz /var/log/messages.gz /var/log/messages.clematis4spiders.gz /var/log/messages.66. Use the command with the compressed file’s name as the argument: [root@bigboy tmp]# zcat /var/log/messages.. and the use of syslog will become increasingly important especially in the troubleshooting of Linux-based firewalls. stat=Deferred: Connection refused by www.gz [root@bigboy tmp]# Viewing the contents of the files still remains easy because the zcat command can quickly output the contents to the screen. relay=www.gz /var/log/messages.gz /var/log/messages..gz extension. delay=2+20:45:44.06Harrison_ch05. which logs all application programming errors generated by some of the popular scripting languages such as PERL and PHP. xdelay=00:00:02..

qxd 2/25/05 9:57 AM Page 100 .06Harrison_ch05.

hasn’t been customized in any way.qxd 2/25/05 9:58 AM Page 101 C H A P T E R 6 Installing RPM Software IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Where to Get Commonly Used RPMs Getting RPMs Using Web-Based FTP Getting RPMs Using Command-Line Anonymous FTP Getting RPMs Using wget Automatic Updates with yum How to Install RPMs Manually How to Install Source RPMs RPM Installation Errors How to List Installed RPMs Listing Files Associated with RPMs Uninstalling RPMs Which RPMs Start at Boot Time? Installing Software Using tar Files Conclusion You’ll frequently need to install additional software on your Linux server that you didn’t think you’d need when you first installed the operating system. Fedora Linux software is primarily available in Red Hat Package Manager (RPM) files. or master program. This is the usual scenario for most departmental servers.07Harrison_ch06. Regular RPM package files are used for installations in which the kernel. This could be because of new business requirements for additional packages or the need to install new administrative tools to make your job easier. Source RPMs are used when the kernel has been customized to add or drop support selectively for various devices or features for the sake of performance or functionality. The 101 .

fedora. but is covered in greater detail in sections to follow. . making RPMs relatively easy to use. Remember that Fedora RPMs may not work on Red Hat operating systems. but also devotes a small section on TAR packages near the end because they are still very important. If you’re new to FTP.qxd 2/25/05 9:58 AM Page 102 102 Installing RPM Software Chapter 6 procedure for installing source RPMs involves recompiling source code to fit the needs of these kernel customizations. or by using the automatic yum utility. Using FTP or Your Web Browser You can download RPMs from Fedora using the following link: http://download.red site. and automated downloads using yum.redhat.com/pub/fedora/linux/core/ hat. This chapter focuses on the RPM format. Software developers who want to use a universally recognizable file format across all flavors of Linux also will make their products available as TAR packages. Both package types use standardized commands for installing the software contained inside.07Harrison_ch06. it’s explained later. Start your search in the /pub/fedora/linux/core/ directory and move down the directory tree. RPMs manually downloaded via a browser.com You also can use FTP to download Fedora from the download. don’t worry. This makes life easier for the software developer who wrote the package as he or she now has only to create a single package to support all types of customizations. WHERE TO GET COMMONLY USED RPMS There are three commonly used sources for RPMs: Fedora distribution CDs. which is the format of choice for Fedora Linux software. RPMs Downloaded from Fedora The two most common ways of getting RPMs from Fedora are by manually using FTP or a Web browser. or the wget utility. Each of these methods is introduced here. but they are never up to date for very long. Instructions are usually contained within a file inside the TAR package to help guide the installation. File Transfer Protocol (FTP) client. We discuss using this method in more detail later. TAR packages are generally more difficult to work with than RPM packages because the archived files within them may or may not need to be compiled and the commands to install the software may vary from package to package. RPMs on Your Installation CDs Installing from your distribution CDs is usually easier than having to download files from a remote Web site.fedora.

A good general purpose source for additional software is www. Save the file to your hard drive. 3.qxd 2/25/05 9:58 AM Page 103 Getting RPMs Using Web-Based FTP 103 Using yum The yum program is installed on Fedora systems by default. This site offers packages for a wide variety of software in various formats for many of the popular versions of Linux.i386. Use your Web browser to go to the Fedora link given earlier. Let’s say you are running Fedora Core 2 and need to download an RPM for the DHCP server application. Go to the pub/fedora/linux/core/2/i386/os/Fedora/RPMS directory. Using the rpmfind Web site To use the rpmfind Web site: . Installing an incorrect version could adversely affect your system.rpmfind. 4. 2. two of the most popular are the Fedora Web site and rpmfind. It has an easy-to-use search function that enables you to locate the software you need quickly. GETTING RPMS USING WEB-BASED FTP There are numerous Web sites from which you can download Fedora RPM software. you can select either of these methods in the following sections to get the software. Most downloads can be done using HTTP by either clicking on the link or using the wget command.net. The methods used to get the software from either site are simple. Note Always remember to select the RPM that matches your version of Linux.net Red Hat and Fedora have only their approved software on their sites. RPMs Downloaded from rpmfind.rpm link. but different enough to be treated separately. It enables you to keep the versions of your software up to date by downloading the required packages with the option of installing them afterward.net.0pl2-6. Using the Fedora Web site To use the Fedora Web site: 1. Click on the dhcp-3. This is discussed in more detail in a later section.07Harrison_ch06. which is explained later.16.

220 Fedora FTP server ready. Type dhcp in the Search box.35.232. Connected to download.com (66.redhat.07Harrison_ch06. First.187. 2. we can use the we have at our disposal: help command to see what options . ftp> pwd 257 “/” ftp> ls 227 Entering Passive Mode (66. Save the file to the Linux box’s hard drive. 6. This way anyone can access the data. we issue the FTP command targeting at the command line: download. or something similar. Click the Search button. Let’s illustrate this with an example of using anonymous FTP to download the SSH package from download.35).fedora.57.35. Using binary mode to transfer files. Name (download.232.com:root): anonymous 331 Please specify the password.com Trying 66.qxd 2/25/05 9:58 AM Page 104 104 Installing RPM Software Chapter 6 Go to the rpmfind link.155) 150 Here comes the directory listing..com.redhat.fedora. Have fun.com [root@bigboy tmp]# ftp download.fedora.. 3.fedora. Click the link. After we’ve logged in. Anonymous FTP enables you to log in and download files from a FTP server using the username anonymous or the shorter username ftp and a password that matches your e-mail address. 5. so you can find them later. It is best to download RPMs to a directory named RPM. 1.redhat. GETTING RPMS USING COMMAND-LINE ANONYMOUS FTP The Web-based method in the last section transparently uses anonymous FTP. 1. The right column has links with the actual names of the rpm files.232.redhat. All transfers are logged. ftp> 2.187. drwxr-xr-x 3 ftp ftp 4096 Oct 29 15:59 pub 226 Directory send OK. Scroll down for the RPM that matches your version of Fedora.187. 4.fedora.redhat. Password: 230 Login successful.

Put a file from the local machine onto the FTP server. Give the directory name on the local machine. Command binary cd dir exit get lcd ls mget mput passive put pwd 3. Same as put.qxd 2/25/05 9:58 AM Page 105 Getting RPMs Using Command-Line Anonymous FTP 105 ftp> help Commands may be abbreviated. I know that the Fedora Core 2 RPMs are located in the pub/fedora/linux/core/2/i386/os/Fedora/RPMS/ directory and will use the cd command to change my directory to there. Same as get.07Harrison_ch06. Change the directory on the local machine. Get a file from the FTP server. but you can use wildcards like *.1 FTP Commands Description Copy files in binary mode. Make the file transfer passive mode.1. We can use the ls command to get a listing of files in this directory: . Commands are: ! $ account append ascii bell binary bye case cd cdup chmod close cr Delete ftp> Debug dir disconnect exit form get glob hash help idle image lcd ls macdef Mdelete mdir mget mkdir mls mode modtime mput newer nmap nlist ntrans open prompt passive proxy sendport put pwd quit quote recv reget rstatus rhelp rename reset restart rmdir runique Send Site size status struct system sunique tenex tick trace type user umask verbose ? The commands you’ll most likely use are listed in Table 6. Bye bye. but you can use wildcards like *. List the names of the files in the current remote directory. By using the Web browsing feature on the Web site ahead of time. Change directory on the FTP server. Same as dir. Table 6.

232.58. we just aborted this download by typing n.35. The hash command prints # hash signs on the screen during the download: ftp> hash Hash mark printing on (1024 bytes/hash mark). .3) 150 Here comes the directory listing. 184281 bytes received in 3.1p234.35. .i386.rpm (184281 bytes). 184281 Oct 28 23:29 openssh-3.07Harrison_ch06.6..rpm local: openssh-3.25) 150 Opening BINARY mode data connection for openssh-3.i386. ftp> ls open* 227 Entering Passive Mode (66..6...i386..6.187. 226 Directory send OK.. . Next we get the file we need and place it in the local directory /usr/rpm.i386.41 secs (53 Kbytes/sec) ftp> Note You can also use wildcards to download the RPMs you need using the mget command.1p234.6...232..1p2-34. -rw-r--r-.58. ftp> mget openssh-3.6.. In the next example.rpm . You’ll be prompted for each of the matching RPM files.i386.qxd 2/25/05 9:58 AM Page 106 106 Installing RPM Software Chapter 6 ftp> cd pub/fedora/linux/core/2/i386/os/Fedora/RPMS/ 250 Directory successfully changed.rpm? n ftp> 5.1p234.rpm remote: openssh-3.. ftp> 4.6. . Finally we use the exit command to leave FTP: ftp> exit 221 Goodbye..187.i386. ftp> lcd /usr/rpm Local directory now /usr/rpm ftp> get openssh-3. ############################################################## ############################################################## ####################################################### 226 File send OK.6* mget openssh-3.rpm 227 Entering Passive Mode (66.1p2-34.1p2-34. root@bigboy tmp]# .

==> PASV .66.0pl2-6. Here is an example downloading a DHCP update from Fedora: [root@bigboy tmp]# wget http://linux.0pl2-6. This provides the added advantage of choosing the most reliable sites in your part of the globe.. done..12K/s ETA 17:38:36 (889..i386. done.i386...rpm --17:38:36-ftp://linux.16. right-click on the desired link...12 KB/s) . Length: 529.. but also all the supporting RPMs..i386.5’ Resolving linux. done.. It has a number of valuable features: ☞ You can configure the URLs of download sites you want to use.. ==> RETR dhcp-3. done. Logged in! ==> SYST . ☞ yum automatically figures out not only the RPM packages that need updating.stanford. Logging in as anonymous .16. ==> PWD .rpm.890 (unauthoritative) 100%[===============================>] 529..16.07Harrison_ch06.18]:21. ==> TYPE I . done.890 00:00 889. ☞ yum makes multiple attempts to download RPMs before failing.. you can then select your SSH/telnet/Linux Terminal login window and type the command wget URL..2.5’ saved [529890] [root@bigboy tmp]# AUTOMATIC UPDATES WITH YUM The yum automatic RPM update program comes as a standard feature of Fedora Core.rpm . Connecting to linux.`dhcp-3.i386.edu[171.edu/pub/mirrors/fedora/linux/core/2/i386/os/ Fedora/RPMS/dhcp-3..0pl2-6.qxd 2/25/05 9:58 AM Page 107 Automatic Updates with yum 107 GETTING RPMS USING WGET The wget command can be used to download files quickly when you already know the URL at which the RPM is located.rpm.i386. ==> CWD /pub/mirrors/fedora/linux/core/2/i386/os/Fedora/RPMS .edu/pub/mirrors/fedora/linux/core/2/i386/os/ Fedora/RPMS/dhcp-3..0pl2-6. It then installs them all.stanford.. You can browse the download site for the RPM you need.rpm => `dhcp-3.edu.16. connected. and select copy shortcut (Windows) or Copy Link Location (Linux). done. . done.16..stanford.0pl2-6. After you have done this.stanford. This is especially convenient if you are logged into your Linux box from another machine running a Web browser.

There will not be RPMs in this subdirectory. ☞ The baseurl URL for [updates-released] would be that of the updates/fedora-version/architecture-type subdirectory of your version of Fedora. Here is a sample sites: yum. Make sure won’t work.2 Section [main] [base] [updates-released] /etc/yum.07Harrison_ch06.log pkgpolicy=newest distroverpkg=fedora-release tolerant=1 exactarch=1 .qxd 2/25/05 9:58 AM Page 108 108 Installing RPM Software Chapter 6 Note Updating packages could cause programs written by you to stop functioning.2: Table 6.com/download/mirrors. or it that of the fedora-version/ version of Fedora.conf file to update Fedora from one of the mirror [main] cachedir=/var/cache/yum debuglevel=2 logfile=/var/log/yum.html Web site to get a listing of alternative download sites.conf file. Browse the sites to find the correct locations of the files: ☞ The baseurl URL for [base] would be architecture-type/os subdirectory of your there is a headers subdirectory here.redhat. Make sure there is a headers subdirectory here. File Format—yum. Configuring yum All the configuration parameters for yum are stored in the The three basic sections are listed in Table 6. There will be RPMs in this subdirectory. which can usually be left alone Contains the URL (ftp:// or http://) of a mirror site that contains the Fedora base configuration RPMs Contains the URL (ftp:// or http://) of a mirror site that contains updated Fedora RPMs The easiest way to determine the exact URLs to use in the baseurl parameters of the [base] and [updates-released] sections of the file is to go to the http://fedora. or it won’t work. especially if they rely on the older version’s features or syntax.conf Description Contains logging and fault-tolerance parameters.

Large RPMs are about twenty-five megabytes in size. and they are updated infrequently. so your network load should be minimal on average with an update once or twice a week per server. and the $basearch variable maps to the base architecture of your server.daily directory that runs as a cron job . yum Note It is probably best to select yum update sites that use HTTP instead of FTP. and Web servers are less likely to have connection limits imposed on them. frequently only selecting the last one in the list: baseurl=url://server1/path/to/files/ url://server2/path/to/files/ url://server3/path/to/files/ Creating Your Own yum Server An obvious advantage of using yum is that you can use it to update a yum server at your office with the same directory structure of the mirror download sites on the Fedora Web site. which often have limits on the number of user logins.$basearch .07Harrison_ch06. The problem is timing.Base baseurl=http://mirrors. The $releasever variable refers to the current version of Fedora Core running on your server. A small desktop PC with about five to six gigabytes of free disk space per distribution should be sufficient to start with for a dedicated small business yum server. yum may act strangely. There is a yum script file in the /etc/cron.com/fedora/core/updates/$releasever/ $basearch/ Note accepts the use of variables in the configuration file.com/fedora/core/$releasever/$basearch/ os/ [updates-released] name=Fedora Core $releasever . There are a number of reasons for this. which is determined automatically. If you use multiple baseurl statements in each section.Released Updates baseurl=http://mirrors.qxd 2/25/05 9:58 AM Page 109 Automatic Updates with yum 109 [base] name=Fedora Core $releasever .xmission. outbound HTTP access to the Internet is often already allowed in offices. FTP firewall rules are more difficult to implement than HTTP. unlike FTP servers. Note You can list multiple URLs in a baseurl statement like this and yum will try them all.xmission.$basearch .

This is set using the gpgcheck variable in the /etc/yum.conf # gpgcheck=1 This is a valuable feature to have but you need to load the checksum files in order for yum to work properly. Please refer to the section titled “Signature Keys” later in the chapter before proceeding.” covers the Apache Web server for HTTP requests. you need to enable the viewing of directory structures so that it will be easy for someone to use his or her Web browser to navigate down the directories and double-check the location of the yum files. “Linux FTP Server Setup.07Harrison_ch06. you can upgrade the server. “Monitoring Server Performance. You can also consider throttling the NIC card of the yum server to 10Mbps as another interim means of reducing the problem. if these measures don’t work.conf file. Before You Start As of Fedora Core version 3. MRTG monitoring is covered in Chapter 22.” discusses Linux FTP servers and Chapter 20. they are disabled: # # File: /etc/yum. When the value is set to 1. “The Apache Web Server. when set to 0. Chapter 15.qxd 2/25/05 9:58 AM Page 110 110 Installing RPM Software Chapter 6 every day at 4:00 a.” When established. and you can use MRTG on the server to get trend data for its network load. yum clients can access the yum server using either FTP or HTTP requests. the yum utility checks the downloaded RPMs against checksum files to help protect against file corruption and malicious forgeries.m. Your yum server could get overwhelmed with simultaneous update requests from all your yum clients. Finally. Keeping Your System Current with yum You can make the installed RPM packages on your system up to date with the latest patches using the yum update command. Note When setting up an HTTP-based yum server. If you need help in setting these up. If the load gets too high. you can then configure all your Fedora servers to use this local yum server for all updates. then checks are done. which will significantly reduce your Internet congestion and the associated bandwidth costs. When used without listing any . For most small businesses/departments this should not be a major concern. you could move this script to another location and schedule it as a cron job for different times for each server.

yum will attempt to update them all. ..i686 Updated: pango 1...rpm samba-client-3.FC1.2-7. . Is this ok [y/N]: y Getting samba-client-3. Here is an example of output with yum updating your system: [root@bigboy tmp]# yum update Gathering header information file(s) from server(s) Server: Fedora Core 2 .65/127 Completing update for samba-client ..66/127 Completing update for binutils .5-4.FC1.. . It is always advisable to use yum after installing Linux to make sure the latest versions of software are installed for the sake of improved security and functionality.90.127/127 Kernel Updated/Installed.07Harrison_ch06.0..14..2. Completing update for XFree86-font-utils .2166... .i386] [update: binutils 2.90. .qxd 2/25/05 9:58 AM Page 111 Automatic Updates with yum 111 package-name packages afterward.2-7..67/127 .4.0..i386.Released Updates Finding updated packages Downloading needed headers Resolving dependencies Dependencies resolved I will do the following: [install: kernel 2.2166. ..i386] .nptl.Base Server: Fedora Core 2 . checking for bootloader Grub found .. .i686] [update: samba-client 3. Success! glibc-common 100 % done 1/127 glibc 100 % done 2/127 Stopping sshd:[ OK ] Starting sshd:[ OK ] bash 100 % done 3/127 mozilla-nspr 100 % done 4/127 sed 100 % done 5/127 .0.nptl.6-4.. 100% |=========================| 128 kB 05:01 . Completing update for pango ..FC1.14.i386 samba-client 3.0.6-4.i386 initscripts [root@bigboy tmp]# ....i386 .2-7.4.0-55.22-1. The yum update command updates only a particular RPM package. Running test transaction: Test transaction complete......i386 binutils 2.making this kernel the default Installed: kernel 2.FC1.0.i386 XFree86-Mesa-libGLU 4. .3.2-7.22-1..0.i386 .

2-11 Downloading Packages: net-snmp-utils-5. not adding again Setting up Install Process Setting up Repo: base repomd.2 MB Running Transaction Test Finished Transaction Test Transaction Test Succeeded Running Transaction Installing: net-snmp-utils 100 % done 1/1 Installed: net-snmp-utils.i386 0:5.2-11. use the yum with the -y switch. In this case [root@bigboy tmp]# yum -y install net-snmp-utils Repository updates-released already added.qxd 2/25/05 9:58 AM Page 112 112 Installing RPM Software Chapter 6 Note If you don’t want to be prompted to install the files. as in this example: [main] exclude=kernel .2-11 set to be installed --> Running transaction check Dependencies Resolved Transaction Listing: Install: net-snmp-utils. Please wait.1.2-11 Complete! [root@bigboy tmp]# 00:48 Remember the following facts about yum: ☞ You can place a list of packages you never want automatically updated in the [main] section.i 100% |===================| 6.1. Kernel RPMs may be one of the first sets to go on this list.xml 100% |=========================| 951 B 00:00 Reading repository metadata in from local files base : ############################################ 2622/2622 primary. not adding again Repository base already added.xml 100% |=========================| 1.1.1 kB 00:00 Setting up Repo: updates-released repomd. ---> Package net-snmp-utils.i386 0:5.i386 0:5.xml.07Harrison_ch06. Example of a yum Package Installation Here is a sample installation of an individual package using the RPM installed is the net-snmp-utils package: yum.gz 100% |=========================| 88 kB 00:00 MD Read : ################################################## 229/229 updates-re: ################################################## 229/229 Resolving Dependencies --> Populating transaction set with selected packages.1. The list must be separated by spaces.

23. such as /tmp. and the other is to install the RPM from some sort of removable media.i386. and the -v qualifier prints verbose status messages while the command is run. the -h qualifier gives a list of hash # characters during the installation.conf command. check the files in the tory. Here is an example of a typical RPM installation command to install the MySQL server package: [root@bigboy tmp]# rpm -Uvh mysql-server-3.58-9.rpm Preparing.qxd 2/25/05 9:58 AM Page 113 How to Install RPMs Manually 113 does its updates using TCP port 80 for http:// update URLs and uses passive FTP for ftp:// update URLs in /etc/yum. ☞ Don’t limit yourself to the default yum.. This will have importance for your firewall rules. ####################### [100%] 1:mysql-server ####################### [100%] [root@bigboy tmp]# Using CD-ROMs The underlying steps to install RPMs from CDs are similar to those used when installing from your hard disk.rpm) into a temporary directory. and then install the RPM: /mnt/cdrom/Fedora/RPMS direc- . Insert the CD-ROM.07Harrison_ch06.conf URLs because they can become overloaded with requests and make yum perform poorly. The cron file is located in /etc/cron. yum ☞ HOW TO INSTALL RPMS MANUALLY There are generally two ways to install RPM files manually. Your RPMs will then be located in the CD-ROM’s Fedora/RPMs subdirectory. ☞ More details on configuring yum can be obtained by running the man yum. The procedure is as follows: 1.daily/. The first method is by using a file previously downloaded to your hard drive.. The next step is to issue the rpm -Uvh command to install the package. ☞ yum runs automatically each day. The main difference is that you have to access your CD-ROM drive by mounting it first to the mnt/cdrom directory.conf. Using Downloaded Files Download the RPMs (which usually have a file extension ending with . such as a CD-ROM drive. The -U qualifier is used for updating an RPM to the latest version.

eject the CD-ROM: [root@bigboy RPMS]# cd /tmp [root@bigboy tmp]# eject cdrom [root@bigboy tmp]# Note You can use the rpm command’s --aid switch to make it search the CD-ROM for any other RPM dependencies and install them automatically.. + umask 022 + cd /usr/src/redhat/BUILD + rm -rf tac_plus-4.tgz + tar -xvvf .rpm [root@bigboy RPMS]# rpm -Uvh filename.src.0..0.src.3 + /usr/bin/gzip -dc /usr/src/redhat/SOURCES/tac_plus-4. HOW TO INSTALL SOURCE RPMS Sometimes the packages you want to install need to be compiled in order to match your kernel version.rpm Installing tac_plus-4.3 + exit 0 [root@bigboy rpm]# .rpm Executing(%prep): /bin/sh -e /var/tmp/rpm-tmp.07Harrison_ch06. ☞ Compiling and installing source RPMs with Fedora can be done simply with the rpmbuild command: [root@bigboy tmp]# rpmbuild --rebuild filename.0.rpm Preparing.src... ####################### [100%] 1: filename ####################### [100%] [root@bigboy RPMS]# 2.. They usually have a file extension ending with (. When finished.rpm)..61594 + umask 022 + cd /usr/src/redhat/BUILD + cd /usr/src/redhat/BUILD + rm -rf tac_plus-4.3. This requires you to use source RPM files: ☞ Download the source RPMs or locate them on your CD collection...3-2.0. .src.3-2. ☞ Run the following commands as root.0. .qxd 2/25/05 9:58 AM Page 114 114 Installing RPM Software Chapter 6 [root@bigboy tmp]# mount /mnt/cdrom [root@bigboy tmp]# cd /mnt/cdrom/Fedora/RPMS [root@bigboy RPMS]# ls filename* filename.rpm ☞ Here is an example in which we install the TACACS plus package: [root@bigboy rpm]# rpmbuild --rebuild tac_plus-4.

23. This section shows you how to recover from some of the most common errors you’ll encounter. RPM INSTALLATION ERRORS Sometimes the installation of RPM software doesn’t go according to plan and you need to take corrective actions. needs to be installed beforehand: [root@bigboy tmp]# rpm -Uvh mysql-server-3. followed by successful installation of perl-DBD-MySQL and mysql-server: [root@bigboy tmp]# rpm -Uvh --nodeps mysql-3. giving Failed dependencies errors.rpm error: Failed dependencies: perl-DBD-MySQL is needed by mysql-3.rpm error: Failed dependencies: libmysqlclient.23.i386. To get around this problem you can run the rpm command with the --nodeps option to disable dependency checks.58-9 mysql = 3.58-9.23. which fails because the mysql MySQL client RPM. Failed Dependencies Sometimes RPM installations will fail.58-9.58-9 [root@bigboy tmp]# MySQL Installing the MySQL client also fails because it requires the package: perl-DBD- [root@bigboy tmp]# rpm -Uvh mysql-3.so. In the next example we install the MySQL client ignoring dependencies. the installation of the perl-DBD-MySQL package fails because it needs the mysql client package.i386. ####################### [100%] 1:mysql ####################### [100%] .58 is needed by mysql-server-3.. You will then have to install the compiled RPMs found in their respective subdirectories as you normally would. on which it depends. which really means that a prerequisite RPM needs to be installed. For example.23.9003-4.i386. if you compiled an i386 architecture version of the RPM it will be placed in the i386 subdirectory.rpm error: Failed dependencies: libmysqlclient. In the next example we’re attempting to install the MySQL database server application.23.9003-4 [root@bigboy tmp]# Strangely enough.10 is needed by mysql-server-3.i386.58-9.so..23.07Harrison_ch06.qxd 2/25/05 9:58 AM Page 115 RPM Installation Errors 115 The compiled RPM file can now be found in one of the architecture subdirectories under /usr/src/redhat/RPMS directory.23.10 is needed by perl-DBD-MySQL-2.58-9 [root@bigboy tmp]# rpm -Uvh perl-DBD-MySQL-2.rpm Preparing.

rpm Preparing.. key ID 4f2a6fd2 error: dhcp-3.0pl2-6.rpm cannot be installed [root@bigboy tmp]# . This can be done using the rpm command as seen in the next example. the RPMs will install with only a warning message. the rpm command can automatically install all the prerequisite RPMs using the --aid option. ####################### [100%] 1:perl-DBD-MySQL ####################### [100%] [root@bigboy tmp]# rpm -Uvh mysql-server-3.23.16..07Harrison_ch06.i386.0pl2-6.0pl2-6..i386.i386.rpm: V3 DSA signature: BAD.58-9. It is a good idea to import both the Red Hat and Fedora keys: [root@bigboy tmp]# rpm --import /usr/share/rhn/RPM-GPG-KEY [root@bigboy tmp]# rpm --import /usr/share/rhn/RPM-GPG-KEY-fedora [root@bigboy tmp]# If you don’t install the keys you get a DSA signature warning that alerts you to the fact that the RPM file might be bogus: [root@bigboy tmp]# rpm -Uvh dhcp-3. If they are not there.i386.rpm: V3 DSA signature: NOKEY.9003-4.16. Signature Keys Fedora digitally signs all its RPM files..rpm Preparing. If the RPM’s digital signature doesn’t match that in the key file. #################################### [100%] 1:dhcp #################################### [100%] [root@bigboy tmp]# It is always good to install the key files.i386.0pl2-6.16. ####################### [100%] 1:mysql-server ####################### [100%] [root@bigboy tmp]# Note If all the installation RPMs are located in the same directory. the rpm installation program also alerts you and fails to install the RPM package at all: [root@bigboy tmp]# rpm -Uvh dhcp-3.0pl2-6. so it’s best to import their public encryption key beforehand so that the RPM installation program will be able to verify the validity of the RPM file.qxd 2/25/05 9:58 AM Page 116 116 Installing RPM Software Chapter 6 [root@bigboy tmp]# rpm -Uvh perl-DBD-MySQL-2.16. key ID 4f2a6fd2 Preparing.rpm warning: dhcp-3. One of the advantages of using the yum facility is that you don’t have to worry about this dependency process as much because the dependency RPMs are always downloaded and installed automatically also..i386.rpm error: dhcp-3.16.i386..

2-7 audiofile-0.3-3 .. HOW TO LIST INSTALLED RPMS The rpm -qa command lists all the packages installed on your system: [root@bigboy tmp]# rpm -qa perl-Storable-1. [root@bigboy tmp]# You can also pipe the output of this command through the grep command if you are interested in only a specific package.4p1-2 [root@bigboy tmp]# Note You could use the rpm -q package-name command to find an installed package because it is much faster than using grep and the -qa switch. [root@bigboy tmp]# rpm -qa | grep -i ssh openssh-server-3.4-9 e2fsprogs-1.2.qxd 2/25/05 9:58 AM Page 117 Listing Files Associated with RPMs 117 Signatures are therefore useful because they help protect you against tampered and otherwise corrupted RPMs being installed. the latter method is probably more suitable..27-9 libstdc++-3. In many cases the installation program doesn’t state the RPM package in which the file can be found. regardless of case (-i means ignore case).14-15 smpeg-gtv-0.0..4p1-2 openssh-askpass-3. It is therefore important to be able to determine the origin of certain files by listing the contents for RPMs in which you suspect the files might reside. If you are not sure of the package name and its capitalization.4..4p1-2 openssh-askpass-gnome-3. .07Harrison_ch06.4p1-2 openssh-clients-3.4p1-2 openssh-3. . LISTING FILES ASSOCIATED WITH RPMS Sometimes you’ll find yourself installing software that terminates with an error requesting the presence of a particular file. In this example we are looking for all packages containing the string ssh in the name.. . but you have to have an exact package match..

.8..2-0. In this example we test to make sure that the NTP package is installed using the-q qualifier.1. /usr/share/doc/ntp-4. and then we use the -ql qualifier to get the file listing: [root@bigboy tmp]# rpm -q ntp ntp-4.qxd 2/25/05 9:58 AM Page 118 118 Installing RPM Software Chapter 6 Listing Files for Already Installed RPMs This can be useful if you have to duplicate a working server that is already in a production environment.rpm /etc/rc.gz /var/lib/dhcp /var/lib/dhcp/dhcpd. /usr/share/man/man8/dhcrelay. In this case you need to know which RPM on the old server contains the file..i386. .htm [root@bigboy tmp]# Listing Files in RPM Files Sometimes you make a guess and download what you think is the RPM with the missing file.conf /etc/ntp/drift /etc/ntp/keys .2/rdebug. You can use the -ql qualifier to list all the files associated with an installed RPM. .....d/dhcrelay /etc/sysconfig/dhcpd /etc/sysconfig/dhcrelay .2/tickadj.1.2/refclock.1.htm /usr/share/doc/ntp-4.1.0pl1-23.07Harrison_ch06. . You can use the -qpl qualifier to list all the files in an RPM archive to make sure before installing it: [root@bigboy updates]# rpm -qpl dhcp-3.. Sometimes the installation of an application fails on the new server due to the lack of a file that resides on the old one.d/init..2 [root@bigboy tmp]# rpm -ql ntp /etc/ntp /etc/ntp..htm /usr/share/doc/ntp-4.d/init..rc1.d/dhcpd /etc/rc.leases [root@bigboy updates]# . ..1.htm /usr/share/doc/ntp-4.2/release.

tar.sourceforge.” which covers the Linux boot process. This is useful when you have a suspicion about the function of a file but are not entirely sure.tar.tar file extension in the filename. . and when they are. The package name given must match that listed in the rpm -qa command because the version of the package is important: [root@bigboy tmp]# rpm -e package-name WHICH RPMS START AT BOOT TIME? The best way to view and configure which RPMs will start at boot time is by using the chkconfig command with the --list switch. The Linux tar command is used to archive files. the MySQL RPM uses the /etc/my.23. For example.tar. the command to use is tar -xzvf filename. and online software libraries such as www.07Harrison_ch06.net.cnf file as its configuration file. not a file named /etc/mysql. which typically have a .cnf mysql-3. The following example confirms the origin of the /etc/my. their file extensions end with .qxd 2/25/05 9:58 AM Page 119 Installing Software Using tar Files 119 Listing the RPM to Which a File Belongs You might need to know the RPM that was used to install a particular file.58-9 [root@zippy tmp]# UNINSTALLING RPMS The rpm -e command erases an installed package.conf as you’d normally expect. “The Linux Boot Process. These files are also frequently compressed in the gzip format. When the archive is compressed.tgz. The commands to extract the data from either type are similar.gz or . A more detailed explanation will be provided in Chapter 7.gz. which can frequently be obtained from the Web sites of software developers. the command to extract the data is tar -xvf filename. INSTALLING SOFTWARE USING TAR FILES Another popular software installation file format is the tar file. When a tar file is uncompressed.cnf file: [root@zippy tmp]# rpm -qf /etc/my.

Here are the initial steps to install tar-based software: 1. or any other type of server—you have to know how to get the software activated when the system reboots.1 linux-software-1.spec [root@bigboy tmp]# This creates a subdirectory with the installation files inside: [root@bigboy tmp]# ls linux-software-1.. are usually easier to use and may be the better method to use for newer Linux users.3. .1]# plugins pluginsREADME REQUIREMENTS subst.. . each with its own customized installation commands. and troubleshooting of many of the most popular Linux server applications used today..3. linux-software-1. RPMs. If the software you install is intended to make your Linux machine permanently run an application—such as a Web server.pm Makefile.gz [root@bigboy tmp]# 2. configuration. CONCLUSION This is just the beginning.3.. Subsequent chapters cover the use. This is covered in Chapter 7..am linux-software.07Harrison_ch06. with the single standardized command format.1/linux-software-plugins.3.3. mail server.in Software installation with tar files can be frustrating. Issue the tar command to extract the files: [root@bigboy tmp]# tar -xvzf linux-software-1.tar.gz linux-software-1. which outlines all the customized steps to install the software.1.3.3.in INSTALL Makefile.1/plugins-scripts/ . testing. Use the cd command to enter the subdirectory and follow the directions listed in the INSTALL and README files: [root@bigboy tmp]# cd linux-software-1.3.1/ linux-software-1. frequently requiring the installation of other supporting tar files.1 [root@bigboy linux-software-1.in NEWS [root@bigboy linux-software-1.spec. The subdirectory usually contains a file called README or INSTALL.tar.qxd 2/25/05 9:58 AM Page 120 120 Installing RPM Software Chapter 6 The tar file installation process usually requires you first to uncompress and extract the contents of the archive in a local subdirectory.1]# ls COPYING install-sh missing depcomp LEGAL mkinstalldirs scripts FAQ lib linux-software.1.spec Helper. which frequently has the same name as the tar file..3. .

or data unit. THE FEDORA BOOT SEQUENCE You might remember when you installed Linux that the installation process prompted you for a list of partitions and the sizes of each in which your filesystems would be placed.08Harrison_ch07. The MBR then needs to know which partitions on the disk have boot loader code specific to their operating systems in their boot sectors and then attempts to boot one of them. When allocating disk space for the partitions. the first sector.qxd 2/25/05 9:58 AM Page 121 C H A P T E R 7 The Linux Boot Process IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ The Fedora Boot Sequence Determining the Default Boot runlevel Getting a GUI Console Get a Basic Text Terminal Without Exiting the GUI System Shutdown and Rebooting How to Set Programs to Run at Each runlevel Conclusion Learning how Linux boots up is critical. When booting from a hard disk. The very first sector of the hard disk is reserved for the same purpose and is called the master boot record (MBR). the PC system BIOS loads and executes the boot loader code in the MBR. which is fairly sophisticated and therefore cannot entirely fit in the 512 bytes of the MBR. 121 . for each partition is always reserved for programmable code used in booting. Read on for the details. Fedora Linux is supplied with the GRUB boot loader. When you have this information you can use it to alter the type of login screen you get as well as which programs start up.

d/rc3.d /etc/rc.gz title Fedora Core (2.d/rc1.xpm.8-1. which does some system checks such as verifying the integrity of the file systems.8-1.1 Example of a grub.d /etc/rc.d/rc2. The structure of this file is discussed further in Chapter 33.6. When Linux begins to boot with its kernel.6.1 shows a typical grub. the second stage boot loader then displays the familiar Fedora branded splash screen that lists all the configured operating system kernels for your choice.d/rc4.d /etc/rc.0) kernel /vmlinuz-2.d /etc/rc.521) root (hd0.1. “Modifying the Kernel to Improve Performance. and starts vital programs needed for the operating system to function properly. it first runs the /sbin/init program.08Harrison_ch07.d /etc/rc. Figure 7.img title Windows 2000 rootnoverify (hd0.6. Table 7. It then inspects the /etc/inittab file to determine Linux’s overall mode of operation or runlevel.8-1.conf file for a system that can boot both Fedora Linux and Windows 2000.qxd 2/25/05 9:58 AM Page 122 122 The Linux Boot Process Chapter 7 The GRUB MBR boot loader merely searches for a special boot partition and loads a second stage boot loader.521.1 Linux Runlevels Directory /etc/rc.d/rc6. A listing of valid runlevels can be seen in Table 7.” default=0 timeout=10 splashimage=(hd0.d/rc5.d Mode/Runlevel 0 1 2 3 4 5 6 Runlevel Description Halt Single-user mode Not used (user-definable) Full multiuser mode (no GUI interface) Not used (user-definable) Full multiuser mode (with GUI interface) Reboot .conf configuration file.d/rc0. which lists all the available operating systems and their booting parameters.1) chainloader +1 Figure 7.0)/grub/splash.conf file.d /etc/rc. When this is complete.521 ro root=LABEL=/ initrd /initrd-2. This then reads the data in the /boot/grub/grub.

..... This makes file management a lot easier. Most Red Hat/Fedora packages place their startup script in the /etc/init. . respectively.. the system boots up with the text interface on the VGA console. .d/rc3. .d directory and place symbolic links (pointers) to this script in the appropriate subdirectory of /etc/rc. ...d/rc6..... K91isdn S10network . .d. the init process then executes startup scripts located in subdirectories of the /etc/rc.unused # 5 . kudzu with a value of 05 will run before wlan with a value of 09........ you get the GUI..d/rc3. each filename in these directories either starts with an S. . Here is a snippet of the file (delete the initdefault line you don’t need): # Default runlevel...Full multiuser mode # 4 . . K75netfs K96pcmcia . K87portmap S09wlan . which means the script should be run when the system is shutting down...d through /etc/rc.. The deletion of a link doesn’t delete the file... if you do not have networking) # 3 .. . K95firstboot S17keytable [root@bigboy tmp]# .Multiuser. . In our example. which signifies the script should be run at startup.d directory: [root@bigboy tmp]# ls /etc/rc. .. The runlevels used by RHS are: # 0 ..Single user mode # 2 ... it won’t be run. without NFS (The same as 3. The number that follows the K or S specifies the position in which the scripts should be run. When set to 3..... . This is explained later... As you can see.d directory... .halt (Do NOT set initdefault to this) # 1 . . . K86nfslock S05kudzu .qxd 2/25/05 9:58 AM Page 123 Determining the Default Boot runlevel 123 Based on the selected runlevel.d..08Harrison_ch07. DETERMINING THE DEFAULT BOOT RUNLEVEL The default boot runlevel is set in the file /etc/inittab with the initdefault variable. Here is a directory listing of the scripts in the /etc/rc.. or a K. K92iptables S12syslog . If a script isn’t there. Fortunately you don’t have to be a scripting/symbolic linking guru to make sure everything works right because Fedora comes with a nifty utility called chkconfig to do it for you.d .X11 # 6 ... ...d/rc0. Scripts used for runlevels 0 to 6 are located in subdirectories /etc/rc. which can then be used for another day. when set to 5..reboot (Do NOT set initdefault to this) . .. in ascending order..

don’t set the initdefault value to 6 or your system will constantly reboot. as mentioned in the preceding section.08Harrison_ch07. ☞ Changing initdefault from 3 to 5. you might want to operate in text mode runlevel 3 most of the time. ☞ Of course. Remember that when you log out you will get the regular text-based console again: [root@bigboy tmp]# startx ☞ Automatic Method: You can have Linux automatically start the X terminal GUI console for every login attempt until your next reboot by using the init command. Servers that double as personal workstations. You will need to edit your initdefault variable in your /etc/inittab file. has an effect on your next reboot. to keep this functionality even after you reboot. . ☞ Most techies tend to boot up with a plain text-based command-line-type interface (runlevel 3). See the following section on how to get a GUI login all the time until the next reboot. or servers that might have to be operated for an extended period of time by relatively nontechnical staff. Remember you can make runlevel 5 permanent even after a reboot by editing the /etc/inittab file. [root@bigboy tmp]# init 5 When the CPU capacity or available memory on your server is low or you want to maximize all system resources.qxd 2/25/05 9:58 AM Page 124 124 The Linux Boot Process Chapter 7 # id:3:initdefault: id:5:initdefault: # Console Text Mode # Console GUI Mode Note the following: ☞ Most home users boot up with a Windows like GUI (runlevel 5). may need to be run at runlevel 5 all the time through the init 5 command. using the GUI only as necessary with the startx command. Setting it to 0 will never allow it to start! GETTING A GUI CONSOLE You have two main options if your system comes up in a text terminal mode on the VGA console and you want to get the GUI: ☞ Manual Method: You can start the X terminal GUI application each time you need it by running the startx command at the VGA console. or vice versa.

This can cause files that are being updated to become corrupted. Using Virtual Consoles By default. which will be outlined in this section.08Harrison_ch07. or worse. You’ll get a new login prompt for each attempt. ☞ Click on Systems Tools and then Terminal. This makes the GUI run as number 7: ☞ You can step through each virtual console session by using the Ctrl-AltF1 through F6 key sequence. you could corrupt the filesystem directory structure. that value is 0. These are defined by the mingetty statements in the /etc/inittab file. Halt/Shut Down the System The init command enables you to change the current runlevel. Using a GUI Terminal Window You can open a GUI-based window with a command prompt inside by doing the following: ☞ Click on the Fedora logo button in the lower-left corner of the screen. Linux has a number of ways to gracefully shut down and reboot your system. only in runlevel 5. Here is an example: [root@bigboy tmp]# init 0 . SYSTEM SHUTDOWN AND REBOOTING It is usually not a good idea to immediately power off your system when you are finished using it. and for a shutdown. ☞ You can get the GUI login with the sequence Ctl-Alt-F7.qxd 2/25/05 9:58 AM Page 125 System Shutdown and Rebooting 125 GET A BASIC TEXT TERMINAL WITHOUT EXITING THE GUI There are a number of ways for you to get a command prompt when running a Linux GUI. or if the GUI is running after launching startx. Linux runs six virtual console or TTY sessions running on the VGA console. The X terminal GUI console creates its own virtual console using the first available TTY that is not controlled by mingetty. This can be important if you need quick access to commands or you are not familiar with the GUI menu option layout.

Another advantage of the shutdown command is that it warns people that the shutdown is going to occur: [root@bigboy tmp]# shutdown -hy 0 Broadcast message from root (pts/0) (Sat Nov The system is going down for system halt NOW! [root@bigboy tmp]# 6 13:15:27 2004): Reboot the System You can also use the entering runlevel 6: init command to reboot the system immediately by [root@bigboy tmp]# init 6 The reboot command has the same effect. It often prompts you as to whether you are sure you want to execute the command. in this case 0 minutes. which can be avoided with the -y switch. but it also sends a warning message to all users: [root@bigboy tmp]# reboot Broadcast message from root (pts/0) (Sat Nov The system is going down for reboot NOW! [root@bigboy tmp]# 6 12:39:31 2004): -r More graceful reboots can be done with the shutdown command using the switch and specifying a delay. and the first argument tells it how long to wait before starting the procedure. which can be used to the same effect. please refer to the man pages for details. The -h switch forces the system to halt. You can also specify shutting down at a specific time of the day.qxd 2/25/05 9:58 AM Page 126 126 The Linux Boot Process Chapter 7 Fedora also has a shutdown command.08Harrison_ch07. which in this case is 10 minutes: [root@bigboy root]# shutdown -ry 10 Broadcast message from root (pts/0) (Sat Nov 6 13:26:39 2004): The system is going DOWN for reboot in 10 minutes! Broadcast message from root (pts/0) (Sat Nov 6 13:27:39 2004): The system is going DOWN for reboot in 9 minutes! .

third-party applications.. Another reason is the recovery of your root password. . It is best to do this from the console.. Broadcast message from root (pts/0) (Sat Nov The system is going down for reboot NOW! 6 13:36:39 2004): Entering Single-user Mode Some activities require you to force the system to log off all users... . bash-2..... this can be done by using the init command to enter runlevel 1... bash-2..qxd 2/25/05 9:58 AM Page 127 System Shutdown and Rebooting 127 . This can be overcome by running the shutdown command with a delay in minutes as the only argument: [root@bigboy tmp]# shutdown 1 Broadcast message from root (pts/0) (Sat Nov 6 13:44:59 2004): The system is going DOWN to maintenance mode in 1 minute! Broadcast message from root (pts/0) (Sat Nov 6 13:45:59 2004): The system is going down to maintenance mode NOW! . because if you do it from a remote terminal session you’ll be logged out: [root@bigboy root]# init 1 . . as mentioned in Chapter 27.05b# .. this gives no prior warning to users. and the shutdown command doesn’t have a single-user mode option. A typical scenario is the addition of a new hard disk.08Harrison_ch07. Switching to Single-user Mode When the system is running normally.05b# Unfortunately.. and networking so that only the systems administrator has access to the system from the VGA console..” or the troubleshooting of a failed boot process. . “Expanding Disk Capacity.

change your password. or the previous systems administrator might move on to a new job without giving it to you. follow these steps: 1. Return to your default runlevel by using the exit command.6. Reverting to Your Default runlevel from Single-user Mode The exit command forces the system to exit runlevel 1 and revert to the default runlevel for the system. the one which allows you to select the various installed kernels on the system. When at the command prompt. When the Linux logo screen appears. Single-user mode assumes the person at the console is the systems administrator root. .08Harrison_ch07.8-1..conf file: grub append> ro root=LABEL=/ 3. . init 3 and init 5) to alter this default behavior: bash-2. Power on your system. The system will continue to boot. Add a 1 to the end of the string: grub append> ro root=LABEL=/ 1 4.” You will get a prompt like this to add boot options to the boot string found in your grub.. 3. 2. 2. 4. Go to the VGA console and press Ctrl-Alt-Del.05b# exit INIT: Entering runlevel: 3 . The system will then shut down in an orderly fashion.. so you don’t have to specify a root username. Reboot the system and enter single-user mode. To recover the root password. type the letter a for “append..521 on an i686 bigboy login: Root Password Recovery Sometimes you might forget the root password.. Press enter.qxd 2/25/05 9:58 AM Page 128 128 The Linux Boot Process Chapter 7 Entering Single-user Mode at the Grub Splash Screen You can enter single-user mode directly after turning on the power to your system: 1.. . but will go straight to the root # prompt without first asking for a username and password. You can also use the init command (for example. Fedora Core release 2 (Tettnang) Kernel 2.

d and the runlevels at which they will be on or off: [root@bigboy tmp]# chkconfig --list keytable 0:off 1:on 2:on 3:on 4:on atd 0:off 1:off 2:off 3:on 4:on syslog 0:off 1:off 2:on 3:on 4:on gpm 0:off 1:off 2:on 3:on 4:on kudzu 0:off 1:off 2:off 3:on 4:on wlan 0:off 1:off 2:on 3:on 4:on sendmail 0:off 1:off 2:off 3:on 4:off netfs 0:off 1:off 2:off 3:on 4:on network 0:off 1:off 2:on 3:on 4:on random 0:off 1:off 2:on 3:on 4:on .. Use chkconfig to Get a Listing of sendmail’s Current Startup Options The chkconfig command can be used with grep to determine the runlevels in which sendmail will run. 5:on 5:on 5:on 5:on 5:on 5:on 5:on 5:on 5:on 5:on 6:off 6:off 6:off 6:off 6:off 6:off 6:off 6:off 6:off 6:off chkconfig Examples You can use chkconfig to change runlevels for particular packages.qxd 2/25/05 9:58 AM Page 129 How to Set Programs to Run at Each runlevel 129 HOW TO SET PROGRAMS TO RUN AT EACH RUNLEVEL As stated earlier. Here we see it will run at levels 3 and 5: [root@bigboy tmp]# chkconfig --list | grep mail sendmail 0:off 1:off 2:off 3:on 4:off 5:on 6:off [root@bigboy tmp]# Switch Off sendmail Starting Up in Levels 3 and 5 The chkconfig command with the --level switch indicates that some action needs to be done at the runlevels entered as its values. You can use this command with the --list switch to get a full listing of packages listed in /etc/init.08Harrison_ch07. Let’s change it so that sendmail doesn’t start up at boot. the chkconfig command can be used to adjust which applications start at each runlevel. The first argument in the command is the package you want to affect. .. Here we see that sendmail starts with a regular startup at runlevel 3 or 5.. and the second defines whether you want it on or off.. In this case we want sendmail not to be started when entering runlevels 3 and 5: [root@bigboy tmp]# chkconfig --level 35 sendmail off [root@bigboy tmp]# .

makes the changes for runlevels 3 and 5 automatically: [root@bigboy tmp]# chkconfig sendmail off chckconfig Because the intention is to shut down also have to stop it from running now: sendmail permanently. Start sendmail again to get it running immediately. respectively.d directory to shut these down permanently? . but some others are less obvious. Some ports are relatively easy to recognize.08Harrison_ch07. but with the on argument. TCP ports 25 and 22 are for mail and SSH. not just after the next reboot. The screen output of the netstat -an command shows a typical case.qxd 2/25/05 9:58 AM Page 130 130 The Linux Boot Process Chapter 7 By not specifying the runlevels with the --level switch. This usually results in your system listening on a variety of unexpected TCP/IP ports that could be used as doors into your system by hackers. we can use chkconfig once more. [root@bigboy tmp]# chkconfig sendmail on [root@bigboy tmp]# chkconfig --list | grep mail sendmail 0:off 1:off 2:off 3:on 4:off 5:on 6:off [root@bigboy tmp]# service sendmail start Starting sendmail: [ OK ] Starting sm-client: [ OK ] [root@bigboy tmp]# Using chkconfig to Improve Security A default Fedora installation automatically starts a number of daemons that you may not necessarily need for a Web server. we might [root@bigboy tmp]# service sendmail stop Shutting down sendmail: [ OK ] Shutting down sm-client: [ OK ] [root@bigboy tmp]# Double-check that sendmail Will Not Start Up We can then use chkconfig to double-check our work: [root@bigboy tmp]# chkconfig --list | grep mail sendmail 0:off 1:off 2:off 3:off 4:off 5:off 6:off [root@bigboy tmp]# Turn On sendmail Again To reactivate sendmail. Should you use the chkconfig command and the scripts in the /etc/init.

xinetd.0.0. ntp. “The NTP Server.0. In the next examples.0:111 0.0:32768 0.. how do you know which startup script is responsible for TCP port 111? The answer is to use the lsof command.0.0.0.0.0.0:* LISTEN tcp 0 0 127.0:* udp 0 0 0.0.0.0:* udp 0 0 0.08Harrison_ch07. If you don’t need any of these three applications.0:32768 0.0.0.0:* LISTEN tcp 0 0 127.0.0:* LISTEN tcp 0 0 127.0.0.0:111 0..” portmap also has many known security flaws that makes it advisable to be run on a secured network.. “Configuring NIS. respectively.0. and XINETD. The portmap daemon is required for the operation of NFS and NIS. .0.0.0:930 0..0. TFTP.0.0.0:* udp 0 0 0.” and it might be required for their operation: [root@ bigboy tmp]# lsof -i tcp:111 COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME portmap 1165 rpc 4u IPv4 2979 TCP *:sunrpc (LISTEN) [root@ bigboy tmp # [root@bigboy COMMAND PID xinetd 1522 (LISTEN) [root@bigboy tmp]# lsof -i tcp:32769 USER FD TYPE DEVICE SIZE NODE NAME root 5u IPv4 2764 TCP probe-001:32769 tmp]# [root@bigboy root]# lsof -i udp:123 COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME ntpd 1321 ntp 4u IPv4 3390 UDP *:ntp .0:631 0.0:* LISTEN tcp 0 0 :::22 :::* LISTEN udp 0 0 0. A number of network applications are reliant on xinetd.0.0.0.0.0.0.1:32769 0.0.” and 30. ..0..0. [root@bigboy tmp]# For example.0:* udp 0 0 0.0.0.0.0:68 0.0.0.. as explained in Chapter 16. which lists all open or actively used files and can be given additional options to extend its scope to include the TCP/IP protocol stack.0:* .qxd 2/25/05 9:58 AM Page 131 How to Set Programs to Run at Each runlevel 131 [root@bigboy tmp]# netstat -an Active Internet connections (servers and established) Proto Recv-Q Send-Q Local Address Foreign Address State tcp 0 0 0.0. “Telnet.0. and UDP port 123 are being used by the portmap. [root@bigboy root]# . “Remote Disk Access with NFS..0.1:25 0.” is required for synchronizing your time with a reliable time source and may be necessary. topics that are covered in Chapters 29. it’s best to shut down portmap permanently.0. we see that TCP ports 111 and 32769.0. and ntp daemons. which is covered in Chapter 24.0:* LISTEN tcp 0 0 0.1:631 0.

d/rpc.d directory. netfs. unless in a test environment. applications listening only on the loopback interface (IP address 127.. Unpredictable results are always undesirable.. Use the Linux man pages. I usually shut down nfs. If you don’t need nfslock. but there is no rpc. Remember to research your options thoroughly before choosing to shut down an application. Full coverage of server and network security is beyond the scope of this book. I keep sendmail running as it is always needed to send and receive mail (see Chapter 21.statd.statd ls: /etc/init. Your needs may be different.qxd 2/25/05 9:58 AM Page 132 132 The Linux Boot Process Chapter 7 In some cases it’s tricky to determine the application based on the results of the lsof command.0.0. password policies.” for details). reference books.08Harrison_ch07. We soon discover that the nfslock daemon uses it. Firewalls. nfslock.d/* /etc/init. Don’t experiment.d/nfslock:[ -x /sbin/rpc. Final Tips on chkconfig Remember the following: ☞ In most cases. depicted as IP address 0. we’ve discovered that TCP port 32768 is being used by rpc. . portmap. but you should always have a security reference guide on hand to guide your final decisions. In the example below. Those listening on all interfaces. .statd to determine which one is responsible for its operation. ☞ Don’t add/remove anything to other runlevels unless you absolutely know what you are doing.statd ] || exit 0 .0.statd file in the /etc/init. physical access restrictions.0. and patch updates need to be considered. The simple solution is to use the grep command to search all the files for the string rpc. and the Internet for information. and cups printing as standard practice on Internet servers. [root@bigboy tmp]# As a rule of thumb. then shut it down permanently: [root@bigboy tmp]# lsof -i tcp:32768 COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME rpc.1) are usually the least susceptible to network attack and probably don’t need to be stopped for network security reasons. you want to modify runlevels 3 and 5 simultaneously and with the same values. Shutting down applications is only a part of server security.statd: No such file or directory [root@bigboy tmp]# grep -i statd /etc/init.d/rpc. are naturally more vulnerable and their continued operation should be dependent on your server’s needs...statd 1178 rpcuser 6u IPv4 2400 TCP *:32768 (LISTEN) [root@bigboy tmp]# ls /etc/init. “Configuring Linux Mail Servers.0.

always try to reboot your system to make sure all the newly installed applications start up correctly. CONCLUSION The topics discussed in this chapter might seem simple.qxd 2/25/05 9:58 AM Page 133 Conclusion 133 ☞ chkconfig doesn’t start the programs in the /etc/init. which was covered in Chapter 5.” they are an essential part of Linux administration that gets frequently overlooked— especially when new software is installed.08Harrison_ch07. Taking the time to configure and test your startup scripts could prevent you from being awakened in the middle of the night while you are on vacation! It is really important.d directory. “Troubleshooting Linux with syslog. but like syslog. Sometimes they start but give errors listed only in the /var/log directory. The commands for starting and stopping the programs covered in this book are covered in each respective chapter. Whenever possible. it just configures them to be started or ignored when the system boots up. .

08Harrison_ch07.qxd 2/25/05 9:58 AM Page 134 .

conf File How to Get DHCP Started DHCP Servers with Multiple NICs Configuring Linux Clients to Use DHCP Configuring Windows Clients to Use DHCP Simple DHCP Troubleshooting Conclusion Normally if you have a cable modem or DSL.” covers how to do this in detail. You can choose to disable the DHCP server feature on your home router and set up a Linux box as the DHCP server. “Installing RPM Software. Chapter 6. If you install a home cable/DSL router between your modem and home network.” DOWNLOAD AND INSTALL THE DHCP PACKAGE Most Red Hat and Fedora Linux software products are available in the RPM format. This chapter covers only the configuration of a DHCP server that provides IP addresses. Downloading and installing RPMs isn’t hard. If you need a refresher. The configuration of a Linux DHCP client that gets its IP address from a DHCP server is covered in Chapter 3.qxd 2/25/05 9:59 AM Page 135 C H A P T E R 8 Configuring the DHCP Server IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Download and Install the DHCP Pacakge The /etc/dhcpd. you get your home PC’s IP address dynamically assigned from your service provider.09Harrison_ch08. your PC will most likely get its IP address at boot time from the home router instead. 135 . “Linux Networking.

1. there must be a subnet section for each interface on your Linux box: ddns-update-style interim ignore client-updates subnet 192.1. The standard DHCP RPM package doesn’t automatically install a /etc/dhcpd.qxd 2/25/05 9:59 AM Page 136 136 Configuring the DHCP Server Chapter 8 When searching for the file. max-lease-time 86400.255. It uses the commands here to configure your network. which you can always use as a guide: /usr/share/doc/dhcp-<version-number>/dhcpd.1rc14-1.conf file to the /etc directory and then you have to edit it.i386.168.168.sample \ /etc/dhcpd.1.168.0p11 RPM file: [root@bigboy tmp]# cp /usr/share/doc/dhcp-3.09Harrison_ch08. # Don’t forward DHCP requests from this # NIC interface to any other NIC # interfaces option ip-forwarding off.conf.conf. .168.0. but you can find a sample copy of dhcpd.0 netmask 255.sample You have to copy the sample dhcpd. Here is the command to do the copying for the version 3.conf file: Most importantly. remember that the DHCP server RPM’s filename usually starts with the word dhcp followed by a version number like this: dhcp-3.1. # Set the default gateway to be used by # the PC clients option routers 192.1. THE /ETC/DHCPD. it reads the file /etc/dhcpd.conf.201 192.conf in the following directory. # Set the amount of time in seconds that # a client may keep the IP address default-lease-time 86400.255.0pl1/dhcpd.conf Here is a quick explanation of the dhcpd.CONF FILE When DHCP starts.conf file.220.rpm.0 { # The range of IP addresses the server # will issue to DHCP enabled PC clients # booting up on the network range 192.

0 netmask 255.09Harrison_ch08. # Set the NTP server to be used by the # DHCP clients option nntp-server 192.255.168.168. # you need to include the following option in the dhcpd.1.qxd 2/25/05 9:59 AM Page 137 The /etc/dhcpd.100.100. # You can also assign specific IP addresses based on the clients’ # ethernet MAC address as follows (Host’s name is “laser-printer”: host laser-printer { hardware ethernet 08:00:2b:4c:59:23.168. fixed-address 192.222.1.168.1.1. Some devices such as network printers default to getting their IP addresses using DHCP. but users need to access them by a fixed IP address to print their documents. These include telling the DHCP clients where to go for services such as finger and IRC. Check the dhcp-options man page after you do your install: [root@bigboy tmp]# man dhcp-options Note The host statement seen in the sample dhcpd.0. # Set the DNS server to be used by the # DHCP clients option domain-name-servers 192. option subnet-mask 255. This can help to reduce systems administration overhead.conf file: option netbios-name-servers 192.conf File 137 # Set the broadcast address and subnet mask # to be used by the DHCP clients option broadcast-address 192. # If you specify a WINS server for your Windows clients.168.255.255.100.255. } } # # List an unused interface here # subnet 192.255.1.0 { } There are many more options statements you can use to configure DHCP.168.conf file can be very useful.2. This statement can be used to always provide specific IP address to DHCP queries from a predefined NIC MAC address. .

If your DHCP server has more than one interface. If you try to do it. Some older Fedora/Red Hat versions of the DHCP server will fail unless there is an existing dhcpd. Use the chkconfig command to get DHCP configured to start at boot: [root@bigboy tmp]# chkconfig dhcpd on 3. It does this by sending a standardized DHCP broadcast request packet to the DHCP server with a source IP address of 255. it sends it to the default gateway. You also can test whether the DHCP process is running with the following command.) Note More information on adding Linux routes and routing may be found in Chapter 3.255. Finally. .255.255. Note You can’t run your DHCP sever on multiple interfaces because you can have only one route to network 255.leases 2.qxd 2/25/05 9:59 AM Page 138 138 Configuring the DHCP Server Chapter 8 HOW TO GET DHCP STARTED To start DHCP: 1.d/dhcpd script to 4. you have to add a route for this 255. you should get a response of plain old process ID numbers: [root@bigboy tmp]# pgrep dhcpd 5.255. Use the command touch /var/lib/dhcp/dhcpd.255. Use the service command to instruct the start/stop/restart DHCP after booting: [root@bigboy tmp]# service dhcpd start [root@bigboy tmp]# service dhcpd stop [root@bigboy tmp]# service dhcpd restart /etc/init.255.leases file. we assume that DHCP requests will be coming in on interface eth0. (In both of the next two examples.09Harrison_ch08. if not.255.leases to create the file if it does not exist: [root@bigboy tmp]# touch /var/lib/dhcp/dhcpd. you’ll discover that DHCP server working on only one interface. always remember to set your PC to get its IP address via DHCP.255. it requests its IP address from the DHCP server. Remember to restart the DHCP process every time you make a change to the conf file for the changes to take effect on the running process. DHCP SERVERS WITH MULTIPLE NICS When a DHCP configured PC boots.255 address so that it knows the interface on which to send the reply.

255 dev eth0 route add If you want this routing state to be maintained after a reboot. To avoid this problem add the route add command to your /etc/rc. CONFIGURING LINUX CLIENTS TO USE DHCP A Linux NIC interface can be configured to obtain its IP address using DHCP with the examples outlined in Chapter 3.qxd 2/25/05 9:59 AM Page 139 Simple DHCP Troubleshooting 139 Temporary Solution You can temporarily add a route to 255. Permanent Solution The new Fedora Linux method of adding static routes doesn’t seem to support sending traffic out an interface that’s not destined for a specific gateway IP address.255. SIMPLE DHCP TROUBLESHOOTING The most common problems with DHCP usually aren’t related to the server. Windows defaults to using DHCP for all its NIC cards so you don’t have to worry about doing any reconfiguration.255. The following sections present simple troubleshooting steps that you can go through to ensure that DHCP is working correctly on your network. after the server is configured correctly there is no need to change any settings and it therefore runs reliably.255.255. The problems usually occur at the DHCP client’s end for a variety of reasons.255 isn’t intended to be relayed to a gateway.255 using the command: [root@bigboy tmp]# route add -host 255.local script.255. use the permanent solution that’s discussed next.09Harrison_ch08. CONFIGURING WINDOWS CLIENTS TO USE DHCP Fortunately. but it should be sent using the MAC address of the DHCP client in the Ethernet frame. The DHCP packet destined for address 255.255. . Please refer to this chapter if you need a quick refresher on how to configure a Linux DHCP client.

254. Creating a Linux DHCP server is straightforward and touches all the major themes in the previous chapters.0 network until the DHCP server becomes available again.255.0. See whether you can ping the DHCP server. . Just remember to make sure that the range of IP addresses issued by all DHCP servers on a network doesn’t overlap because it could possibly cause unexpected errors.qxd 2/25/05 9:59 AM Page 140 140 Configuring the DHCP Server Chapter 8 DHCP Clients Obtaining 169.254. but before we do. but it is an interesting project to try. CONCLUSION In most home-based networks.255. we’ll do a quick refresher on how to create the Linux users who’ll be using many of the applications outlined in the rest of the book.0. You might want to disable the router/firewall’s DHCP server capabilities to experiment with your new Linux server. Pay special attention to your 255. Give your DHCP client a static IP address from the same range that the DHCP server is supposed to provide. Ensure that your DHCP server is configured correctly and use the pgrep command discussed earlier to make sure the DHCP process is running.255 route. If you cannot. a DHCP server isn’t necessary because the DSL router/firewall usually has DHCP capabilities. 2. A DHCP server may be invaluable in an office environment where the time and cost of getting a network engineer to get the work done may make it simpler for Linux systems administrators to do it by themselves. especially if your DHCP server has multiple interfaces. This is frequently referred to as Automatic Private IP Addressing (APIPA). Now it’s time to try something harder.0 Addresses Whenever Microsoft DHCP clients are unable to contact their DHCP server they default to selecting their own IP address from the 169. double-check your cabling and your NICs.09Harrison_ch08. Here are some steps you can go through to resolve the problem: 1.

This user has a user ID of 0. but is a good memory refresher. You can use the command man useradd to get the help pages on adding users with the useradd command or the man usermod to become more familiar with modifying users with the usermod command. You need to log in as user root to add new users to your Linux server.10Harrison_ch09. which is universally identified by Linux applications as belonging to a user with supreme privileges.qxd 2/25/05 10:01 AM Page 141 C H A P T E R 9 Linux Users and sudo IN THIS CHAPTER ☞ Adding Users ☞ Using sudo ☞ Conclusion Before we proceed. it would be best to cover some basic user administration topics that will be very useful in later chapters. Here you’ll find some simple examples to provide a foundation for future chapters. It is not intended to be comprehensive. ADDING USERS One of the most important activities in administering a Linux box is the addition of users. Who Is the Super User? The super user with unrestricted access to all system resources and files in Linux is the user named root. 141 .

When each new user first logs in. Add the Linux groups to your server: [root@bigboy tmp]# groupadd parents [root@bigboy tmp]# groupadd children [root@bigboy tmp]# groupadd soho 3. this is also known as the User Private Group Scheme.10Harrison_ch09. children. and soho: Parents Paul Jane Children Alice Derek Soho Accounts Sales 2. The directory name will be the same as the user’s username: [root@bigboy tmp]# ll /home drwxr-xr-x 2 root root lost+found drwx-----2 accounts soho drwx-----2 alice children drwx-----2 derek children drwx-----2 jane parents drwx-----2 paul parents drwx-----2 sales soho sales[root@bigboy tmp]# 12288 Jul 24 20:04 1024 1024 1024 1024 1024 1024 Jul Jul Jul Jul Jul Jul 24 24 24 24 24 24 20:33 20:33 20:33 20:33 20:33 20:33 accounts alice derek jane paul How to Change Passwords You need to create passwords for each account. Arrange your list of users into groups by function. 4. they are prompted for their new permanent password. This is done with the passwd command. Each user’s personal directory is placed in the /home directory.qxd 2/25/05 10:01 AM Page 142 142 Linux Users and sudo Chapter 9 How to Add Users Adding users takes some planning. Add the Linux users and assign them to their respective groups: [root@bigboy [root@bigboy [root@bigboy [root@bigboy [root@bigboy [root@bigboy tmp]# tmp]# tmp]# tmp]# tmp]# tmp]# useradd useradd useradd useradd useradd useradd -g -g -g -g -g -g parents paul parents jane children derek children alice soho accounts soho sales If you don’t specify the group with the -g. read through these steps before starting: 1. You are prompted once for your old password and twice for the new one: ☞ User root changing the password for user paul: . In this example there are three groups: parents. Red Hat/Fedora Linux creates a group with the same name as the user you just created.

Here is how unprivileged user paul would change his own password: [paul@bigboy paul]$ passwd Changing password for paul Old password: your current password Enter the new password (minimum of 5. the username: [root@bigboy tmp]# userdel paul There is also an optional -r switch that additionally removes all the contents of the user’s home directory. maximum of 8 characters) Please use a combination of upper and lower case letters and numbers. The command has a single argument. [root@bigboy root]# ☞ Users might want to change their passwords at a future date. New password: your new password Retype new password: your new password passwd: all authentication tokens updated successfully. Use this option with care. The data in a user’s directory can often be important even after the person has left your company: [root@bigboy tmp]# userdel -r paul How to Tell the Groups to Which a User Belongs Use the groups command with the username as the argument: [root@bigboy root]# groups paul paul : parents [root@bigboy root]# . [paul@bigboy paul]$ How to Delete Users The userdel command is used to remove the user’s record from the /etc/passwd and /etc/shadow used in the login process.qxd 2/25/05 10:01 AM Page 143 Adding Users 143 [root@bigboy root]# passwd paul Changing password for user paul. New password: your new password Re-enter new password: your new password Password changed.10Harrison_ch09.

the command fails: [bob@bigboy bob]$ more /etc/sudoers /etc/sudoers: Permission denied [bob@bigboy bob]$ . The commands can be run as user root or as any other user defined in the /etc/sudoers configuration file. it is normally not a good idea for them all to use the root account. The sudo utility was designed to overcome this difficulty. The sudo utility allows users defined in the /etc/sudoers configuration file to have temporary access to run commands they would not normally be able to due to file permission restrictions.10Harrison_ch09. when. USING SUDO If a server needs to be administered by a number of people.txt [root@bigboy tmp]# You can also use the chown command with the -r switch for it to do recursive searches down into directories to change permissions. and where if everyone logs in with the same credentials. This is because it becomes difficult to determine exactly who did what.txt [root@bigboy tmp]# ll test. You may run other privileged commands using sudo within a fiveminute period without being re-prompted for a password.txt [root@bigboy tmp]# chown testuser:users test. you will be prompted for your regular password before it is executed.txt from being owned by user root and group root to being owned by user testuser in the group users: [root@bigboy tmp]# ll test. Without sudo. The privileged command you want to run must first begin with the word sudo followed by the command’s regular syntax. All commands run as sudo are logged in the log file /var/log/messages.1 testuser users 0 Nov 17 22:14 test. When running the command with the sudo prefix.1 root root 0 Nov 17 22:14 test. user bob attempts to view the contents of the /etc/sudoers file. Example of a User Using sudo In this example.txt -rw-r--r-. which is an action that normally requires privileged access. The first argument is the desired username and group ownership for the file separated by a colon (:) followed by the filename. In the next example we change the ownership of the file named text.qxd 2/25/05 10:01 AM Page 144 144 Linux Users and sudo Chapter 9 How to Change the Ownership of a File You can change the ownership of a file with the chown command.txt -rw-r--r-.

visudo uses the same commands as the vi text editor. The visudo Command The visudo command is a text editor that mimics the vi editor that is used to edit the /etc/sudoers configuration file.. . There are a number of guidelines that need to be followed when editing it with visudo.10Harrison_ch09. It is not recommended that you use any other editor to modify your sudo parameters because the sudoers file isn’t located in the same directory on all versions of Linux. The visudo command must run as user root and should have no arguments: [root@bigboy tmp]# visudo The /etc/sudoers File The /etc/sudoers file contains all the configuration and permission parameters needed for sudo to work. separated by commas.. General /etc/sudoers Guidelines The /etc/sudoers file has the general format shown in Table 9... Table 9. ☞ You can have multiple usernames per line. The Linux user group users would be represented by %users.1 Format of the /etc/sudoers File General sudoers File Record Format usernames/group servername = (usernames command can be run as) command There are some general guidelines when editing this file: ☞ Groups are the same as user groups and are differentiated from regular users by a % at the beginning. [bob@bigboy bob]$ The details of configuring and installing sections. . which eliminates the need to anything more in this regard. Downloading and Installing the sudo Package sudo are covered in later Fortunately the package is installed by default by Red Hat/Fedora.1.qxd 2/25/05 10:01 AM Page 145 Using sudo 145 Bob tries again using sudo and his regular user password and is successful: [bob@bigboy bob]$ sudo more /etc/sudoers Password: .

Granting All Access to Specific Users You can grant users this sudoers entry: bob. In most cases. ☞ The keyword ALL can mean all usernames. ☞ sudo assumes that the sudoers file will be used networkwide. /usr/sbin.qxd 2/25/05 10:01 AM Page 146 146 Linux Users and sudo Chapter 9 ☞ Multiple commands also can be separated by commas. ☞ If you run out of space on a line. the file is used by only one server and the keyword ALL suffices for the servername. %operator ALL= /sbin/. you can end it with a back slash (\) and continue on the next line. . thereby bypassing the command logging features of sudo.pl Notice also that the lack of any username entries within parentheses () after the = sign prevents the users from running the commands automatically masquerading as another user.1. which will be using it in the servername position in Table 9. This is explained further in the next example. Spaces are considered part of the command. commands.10Harrison_ch09. Notice how the trailing slash (/) is required to specify a directory location: peter. ☞ The NOPASSWD keyword provides access without prompting you for your password. plus the privilege of running the command /usr/local/apps/check. Granting Access to Specific Users to Specific Files This entry allows user peter and all the members of the group operator to gain access to all the program files in the /sbin and /usr/sbin directories. and servers. groups. /usr/local/apps/check. The example on using aliases in the sudoers file shows how to eliminate this problem. Simple /etc/sudoers Examples This section presents simple examples of how to do many commonly required tasks using the sudo utility.pl. and therefore offers the option to specify the names of servers. bunny bob and bunny full access to all privileged commands with ALL=(ALL) ALL This is generally not a good idea because this allows bob and bunny to use the su command to grant themselves permanent root privileges.

The sudoers file allows users to be grouped according to function with the group and then be assigned a nickname or alias. %operator ALL = !/usr/bin/su. Users ADMINS are then denied the option of running any SHELLS commands and su: Cmnd_Alias SHELLS = /usr/bin/sh. !SHELLS . In the next example. but first you have to be granted this privilege in the sudoers file. All the command shell programs are then assigned to the command alias SHELLS. bob. users peter. This feature can be convenient for programmers who sometimes need to kill processes related to projects they are working on.10Harrison_ch09. requiring peter to stop it with the /bin/kill. bob. /usr/bin/kill /usr/bin/pkill User peter is allowed to stop the monthend process with this command: [peter@bigboy peter]# sudo -u accounts pkill monthend Granting Access Without Needing Passwords This example allows all users in the group operator to execute all the commands in the /sbin directory without the need for entering a password. /usr/bin/kill. \ /usr/bin/rsh. /usr/bin/csh. /usr/local/bin/tcsh. programmer peter is on the team developing a financial package that runs a program called monthend as user accounts. /usr/local/bin/zsh User_Alias ADMINS ADMINS = peter. The sudoers entry would look like this: peter ALL=(accounts) /bin/kill.qxd 2/25/05 10:01 AM Page 147 Using sudo 147 Granting Access to Specific Files as Another User The sudo -u entry enables you to execute a command as if you were another user. Groupings of commands can also be assigned aliases. bunny. For example. and bunny and all the users in the operator group are made part of the user alias ADMINS. This has the added advantage of being more convenient to the user: %operator ALL= NOPASSWD: /sbin/ Using Aliases in the sudoers File Sometimes you need to assign random groupings of users from various departments very similar sets of privileges. which is used throughout the rest of the file. or /usr/bin/pkill commands but only as user accounts. From time to time the application fails. \ /usr/bin/ksh.

All the sudo log entries have the word sudo in them. [root@bigboy tmp]# grep sudo /var/log/messages Nov 18 22:50:30 bigboy sudo(pam_unix)[26812]: authentication failure. USER=root . Using syslog to Track All sudo Commands /etc/sudoers file options by issuing the All sudo commands are logged in the log file /var/log/messages. These accounts aren’t always created automatically. the sudo utility provides a means of dispersing the responsibility of systems management to multiple users. logname=bob uid=0 euid=0 tty=pts/0 ruser= rhost= user=bob Nov 18 22:51:25 bigboy sudo: bob : TTY=pts/0 . PWD=/etc . so you can easily get a thread of commands used by using the grep command to filter the output accordingly. It doesn’t prevent them from copying the files to other locations to be run. This makes sudo a valuable part of any company’s server administration and security policy. but the restrictions can be enforced only as part of the company’s overall security policy. which can be very helpful in determining how user error might have contributed to a problem. the MySQL database application runs as user mysql and the Apache Web server application runs as user apache. The advantage of this is that it helps to create an audit trail. COMMAND=/bin/more sudoers [root@bigboy tmp]# CONCLUSION It is important to know how to add users. or enter command shells that bypass sudo’s command logging. immediately followed by the successful execution of the command /bin/more sudoers. . Here is sample output from a user bob failing to enter his correct sudo password when issuing a command. not just so they can log in to our system.10Harrison_ch09. Finally. Other Examples You can view a comprehensive list of command man sudoers.qxd 2/25/05 10:01 AM Page 148 148 Linux Users and sudo Chapter 9 This attempts to ensure that users don’t permanently su to become root. especially if the software is installed using TAR files. Most server-based applications usually run via a dedicated unprivileged user account—for example. You can even give some groups of users only partial access to privileged commands depending on their roles in the organization.

This chapter describes how you can make your Linux box into a Windows Primary Domain Controller (PDC) or a server for a Windows Workgroup. but this simple explanation should be enough: 149 . such as printers. Linux. What’s the difference between a PDC and Windows Workgroup member? A detailed description is beyond the scope of this chapter. with Windows boxes. Either configuration will allow everyone at home to have: ☞ Their own logins on all the home Windows boxes while having their files on the Linux box appear to be located on a new Windows drive ☞ Shared access to printers on the Linux box ☞ Shared files accessible only to members of their Linux user group.qxd 2/25/05 10:02 AM Page 149 C H A P T E R 10 Windows. and Samba IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Download and Install Packages How to Get Samba Started The Samba Configuration File How SWAT Makes Samba Simpler Creating a Starter Configuration Samba Passwords How to Create a Samba PDC Administrator User How to Add Workstations to Your Samba Domain How to Add Users to Your Samba Domain Domain Groups and Samba How to Delete Users from Your Samba Domain How to Modify Samba Passwords Conclusion Samba is a suite of utilities that allows your Linux box to share files and other resources.11Harrison_ch10.

each PC stores the usernames and passwords locally so that they are unique for each PC. as in samba-client3.” covers how to do this in detail.0. You can create your own smb. “Installing RPM Software. test the validity of the changes using the testparm utility outlined in Chapter 12. you will lose all the comments you inserted with the text editor. . ☞ In a Windows Workgroup.conf with a text editor then subsequently use SWAT to edit the file. Samba is comprised of a suite of RPMs that come on the Fedora CDs. If you need a refresher. Samba domains and Linux share the same usernames so you can log into the Samba-based Windows domain using your Linux password and immediately gain access to files in your Linux user’s home directory.” and when you are satisfied with your changes. which is easier. and Samba Chapter 10 ☞ A PDC stores the login information in a central database on its hard drive. DOWNLOAD AND INSTALL PACKAGES Most Red Hat and Fedora Linux software products are available in the RPM format. you can make your Samba and Linux passwords different. Samba mimics a Windows PDC in almost every way needed for simple file sharing. By default. and with every client request. Downloading and installing RPMs isn’t hard. For added security. however. This chapter will cover only the much more popular PDC methodology used at home.11Harrison_ch10. Tip As your smb. you’re better off editing a copy of it if you decide not to use SWAT. copy the file back to its original location.conf to determine its various modes of operation. I’ll explain how to use both SWAT and a text editor to configure Samba later in this chapter. Chapter 6. that if you create /etc/samba/smb. This allows each user to have a universal username and password when logging in from all PCs on the network.qxd 2/25/05 10:02 AM Page 150 150 Windows. remember that the RPM’s filename usually starts with the RPM name followed by a version number. Linux. The files are named ☞ ☞ ☞ ☞ samba samba-common samba-client samba-swat When searching for the file.conf using a text editor or the Web-based SWAT utility.conf is constantly being accessed. the Samba daemon reads the configuration file /etc/samba/smb. Linux functionality doesn’t disappear when you do this.0-15. When it starts up.i386. “Samba Security and Troubleshooting. After completing your modifications. Keep in mind.

) A share for storing domain logon information such as favorites and desktop icons. and restart Samba after boot time using the tialization script as in the examples below: [root@bigboy tmp]# service smb start [root@bigboy tmp]# service smb stop [root@bigboy tmp]# service smb restart smb ini- Note Unlike many Linux packages. stop. Used for configuring printers.11Harrison_ch10. It is split into five major sections. A share for storing logon scripts.qxd 2/25/05 10:02 AM Page 151 The Samba Configuration File 151 HOW TO GET SAMBA STARTED You can configure Samba to start at boot time using the chkconfig command: [root@bigboy tmp]# chkconfig smb on You can start. you should get a response of plain old process ID numbers: [root@bigboy tmp]# pgrep smb THE SAMBA CONFIGURATION FILE The /etc/samba/smb. which Table 10.) Section [global] [printers] [homes] [netlogon] [profile] You can edit this file by hand or more simply through Samba’s SWAT Web interface. .conf Description General Samba configuration parameters. as it is read after the receipt of every client request. (Not created by default. Defines treatment of user logins. You can test whether the smb process is running with the pgrep command.1 File Format of smb.conf file is the main configuration file you’ll need to edit. (Not created by default. Samba does not need to be restarted after changes have been made to its configuration file.1 outlines. Table 10.

conf file into which you fill in the desired parameters. Linux. Basic SWAT Setup You must always remember that SWAT edits the smb. Figure 10. For ease of use.conf file but also strips out any comments you may have manually entered into it beforehand. each parameter box has its own online help.conf /etc/samba/smb.1 shows the main SWAT login screen. as in [root@bigboy tmp]# cp /etc/samba/smb. Each SWAT screen is actually a form that covers a separate section of the smb.original . and Samba Chapter 10 HOW SWAT MAKES SAMBA SIMPLER SWAT.original.11Harrison_ch10.conf file has many worthwhile comments in it. enables you to configure your smb. you could save the original file with the name /etc/samba/smb.conf. you should save a copy as a reference before proceeding with SWAT. Figure 10.conf file without needing to remember all the formatting.qxd 2/25/05 10:02 AM Page 152 152 Windows.conf. For example.1 Samba SWAT main menu. Samba’s Web-based configuration tool. The original Samba smb.

168.0.100 http://192. Notice that there are no commas between the entries.168.100:901.168.1. as you will see later. because it eliminates these comments.1. .” via a configuration file named /etc/xinetd. ☞ The default configuration allows SWAT Web access from the VGA console only as user root on port 901 with the Linux root password.168.1.11Harrison_ch10.d/swat. using SWAT requires some understanding of the smb. SWAT doesn’t encrypt your login password.168. The enabling and disabling.0. Because this could be a security concern in a corporate environment.0:901 in your browser to get the login screen. which is covered in Chapter 16.3 using the URL Remember that most firewalls don’t allow TCP port 901 through their filters. only_from = localhost 192.3 and localhost. and xinetd. ☞ The disable parameter must be set to no to accept connections.3 bigboy IP with address 192. Here is a sample: service swat { port socket_type protocol wait user server log_on_failure disable only_from } = 901 = stream = tcp = no = root = /usr/sbin/swat += USERID = no = localhost The file’s formatting is fairly easy to understand. You may have to adjust your rules for this traffic to pass.qxd 2/25/05 10:02 AM Page 153 How SWAT Makes Samba Simpler 153 As you can see. starting and stopping of SWAT is controlled by xinetd.1. especially as there are only two entries of interest. Therefore in this case you can also configure Samba on your Linux server from PC 192. “Telnet. This means you’ll have to enter http://127. Here’s an example of an entry to allow connections only from 192.1. TFTP. This can automatically be switched between yes and no. You can make SWAT accessible from other servers by adding IP address entries to the only_from parameter of the SWAT configuration file. Become familiar with the most important options in this file before proceeding with SWAT.conf file parameters. you might want to create a Samba administrator user that has no root privileges or only enable SWAT access from the GUI console or localhost.

[root@bigboy certs]# The resulting certificate has only a 365 day lifetime. Here is how it’s done.) Create an stunnel User You can create a stunnel user via the useradd command: [root@smallfry tmp]# useradd stunnel Create the Certificates From the /usr/share/ssl/certs directory.11Harrison_ch10. the chkconfig command automatically modifies the disable field accordingly in the configuration file and activates the change.pem .” for more details on encryption methods... . [root@bigboy tmp]# cd /usr/share/ssl/certs [root@bigboy certs]# make stunnel. When running SWAT in the unsecured mode above you should take the added precaution of using it from the Linux console whenever possible.qxd 2/25/05 10:02 AM Page 154 154 Windows. Common Name (eg.16. “Miscellaneous Linux Topics. Remember to repeat this process next year. Linux. but make sure you use the server’s IP address when prompted for your server’s Common Name or hostname.. and Samba Chapter 10 Controlling SWAT As with all xinetd-controlled applications. you can create the encryption key certificate using the make command..200 . Use all the defaults when prompted. To activate SWAT use [root@bigboy tmp] chkconfig swat on To deactivate SWAT use [root@bigboy tmp] chkconfig swat off Encrypting SWAT By default SWAT is configured via an unencrypted Web link using the Linux root account.1. your name or your server’s hostname) []: 172. (Please refer to the VPN section of Appendix I. You can configure SWAT to work only with securely encrypted HTTP (HTTPS) versus the regular HTTP method shown above.

conf Configuration File You can configure the stunnel application to: ☞ Intercept encrypted SSL traffic received on any TCP port ☞ Decrypt this traffic ☞ Funnel the unencrypted data to any application listening on another port. For example. Client = no cert = /usr/share/ssl/certs/stunnel. [swat] accept = 901 connect = 902 . Use the [root@bigboy certs]# chmod 640 stunnel.qxd 2/25/05 10:02 AM Page 155 How SWAT Makes Samba Simpler 155 Modify Certificate File Permissions The certificate needs to only be read by chmod and chgrp commands to do this.1 root stunnel 1991 Jul 31 21:50 stunnel.pem key = /usr/share/ssl/certs/stunnel.pem [root@bigboy certs]# ll /usr/share/ssl/certs -rw-r----. you can configure the /etc/stunnel/stunnel.pem # Accept SSL connections on port 901 and funnel it to # port 902 for swat. root and the stunnel user. Here’s how: # Configure stunnel to run as user “stunnel” placing temporary # files in the /home/stunnel/ directory chroot = /home/stunnel/ pid = /stunnel.pid setuid = stunnel setgid = stunnel # Log all stunnel messages to /var/log/messages debug = 7 output = /var/log/messages # Define where the SSL certificates can be found.pem [root@bigboy certs]# Create an /etc/stunnel/stunnel.conf file to intercept SSL traffic on the SWAT port 901 and funnel it decrypted to a SWAT daemon running on port 902.pem [root@bigboy certs]# chgrp stunnel stunnel.11Harrison_ch10.

copy the swat file and name it swat-stunnel. it ignores the configuration file. [root@bigboy certs]# cd /etc/xinetd. If the daemon doesn’t find a match.1 = root = /usr/sbin/swat += USERID = no = 127.qxd 2/25/05 10:02 AM Page 156 156 Windows.d]# stunnel . Also make sure that the service is set to swat-stunnel. You’ll also need to shutdown regular SWAT beforehand.0.1 Edit the /etc/services File to Create a Secure SWAT Entry The xinetd daemon searches the /etc/services file for ports and services that match those listed in each configuration file in the /etc/xinetd.11Harrison_ch10.d]# cp swat swat-stunnel Your new swat-tunnel file should look like this: service swat-stunnel { port socket_type wait only_from user server log_on_failure disable bind } = 902 = stream = no = 127.0.d File for Secure SWAT To start. You now have to edit /etc/services to include your new swat-stunnel file: swat-stunnel 902/tcp # Samba Web Administration Tool (Stunnel) Activate swat-stunnel You can then start the new swat-stunnel application with the mand.0.d]# chkconfig swat-stunnel on Start stunnel chkconfig com- Now start stunnel for the encryption to take place: [root@bigboy xinetd. Configure the new file to be enabled.d]# chkconfig swat off [root@bigboy xinetd. Linux.d directory. and Samba Chapter 10 Create a New /etc/xinetd. [root@bigboy xinetd. listening on port 902 and accepting connections only from localhost.0.d [root@bigboy xinetd.

d]# netstat -an . [root@bigboy xinetd. After this.0:* LISTEN LISTEN Test the Secure SWAT Login Point your browser to the Samba server to make an HTTPS connection on port 901: https://server-ip-address:901/ You will be prompted for the Linux root user username and password.0:* 0.0. you may get a cryptonet error when starting stunnel.0...0.0. . as in: Unable to open “/dev/cryptonet” This is caused by an incompatibility with the hwcrypto RPM used for hardware.qxd 2/25/05 10:02 AM Page 157 How SWAT Makes Samba Simpler 157 In Fedora Core 2.d]# 0.0..d directory and needs to be terminated manually using the pkill command...0. Unfortunately stunnel doesn’t have a startup script in the /etc/init. You need to uninstall hwcrypto to get stunnel to work correctly: [root@bigboy xinetd.11Harrison_ch10.not software-based encryption. [root@bigboy xinetd.0. and start stunnel again.. restart xinetd. tcp 0 0 0. [root@bigboy xinetd.0. . stunnel should begin to function correctly..d]# pkill stunnel [root@bigboy xinetd. Expect a delay of about 60 to 75 seconds with each login.d]# stunnel Test Secure SWAT Your Samba server should now be listening on both port 901 and 902 as shown in the netstat -an command that follows.0:901 tcp 0 0 127. The server will accept remote connections on port 901 only.:902 ..d]# rpm -e hwcrypto You then have to stop stunnel. .

there is no stunnel script in the /etc/init. As mentioned previously. and printers on the network.2 explains the parameters you need to set to create a PDC. Linux.conf file. so I’ll walk you through what you’ll find in each.” to isolate connectivity issues between the SWAT client and Samba server on TCP port 901 amongst other things.conf file. You have to use the pkill command to stop it and the stunnel command to start it again: [root@bigboy tmp]# pkill stunnel . CREATING A STARTER CONFIGURATION The easiest way to configure a Samba server to be the PDC for a small network is by using SWAT. The [global] Section The [global] section governs the general Samba settings.qxd 2/25/05 10:02 AM Page 158 158 Windows.11Harrison_ch10. Makes the PDC act as the central store for the names of all Windows clients. You’ll need to edit the various sections of the smb. “Simple Network Troubleshooting. How to Make SWAT Changes Immediate SWAT immediately changes the functioning of Samba whenever you commit your changes through the Web GUI. and Samba Chapter 10 Troubleshooting Secure SWAT Sometimes you’ll make mistakes in the stunnel.2 Minimum Settings for the [global] Section of smb. Also known as a local master browser. The local master browsers register themselves with the domain master to learn about resources on other networks. stunnel Make sure the file permissions and ownership on the stunnel.pem file are correct and that SWAT is always permanently off but swat-stunnel is permanently on.local file for this to occur. Table 10. Table 10. You can also refer to Chapter 4. servers. Very helpful when you need to browse your local network for resources.d directory to easily stop and restart it. Unfortunately. You will have to add the command stunnel in the /etc/rc. stunnel doesn't start automatically when the system reboots because it doesn't have its own startup script. but changes to this file take effect only after stunnel has been restarted. Parameter domain logons preferred master domain master Yes .conf Value Yes Yes Description Tells Samba to become the PDC. Tells Samba to become the master browser across multiple networks all over the domain.

conf file should resemble the example below when you’re finished. You can view the contents of the configuration file by logging into the Samba server via a command prompt and using cat /etc/samba/smb. 3. User Makes domain logins query the Samba password database located on the Samba server itself. Here’s how to set the values using SWAT: 1. even though you may see them in SWAT. 4. and click on the [global] section. For example. However the defaults may not be to your liking. Make your changes. and click on the Commit Changes button when finished. “homenet” The name of the Windows domain you’ll create. and some of the [global] parameters mentioned previously will be set to Auto. In other words. [global] workgroup = HOMENET time server = Yes domain logons = Yes os level = 65 preferred master = Yes domain masterc = Yes Be aware: Security [eq] user and WINS support [eq] yes are default settings for Samba. 2. it keeps track of the IP addresses of all the domain’s servers and clients. and they may not show up in your smb. I’ve decided to use homenet. Log into SWAT. The name you select is your choice. the default domain is MYGROUP. Using the SWAT Wizard The SWAT utility has a Wizard button that can be used to configure your server as a PDC quickly. . Yes Allows the Samba server to provide name services for the network.conf file.conf to verify your changes as you do them. Yes Lets the Samba server provide time updates for the domain’s clients. Click the Advanced button to see all the options.qxd 2/25/05 10:02 AM Page 159 Creating a Starter Configuration 159 Parameter Os level Value 65 Description wins support time server domain security Sets the priority the Samba server should use when negotiating to become the PDC with other Windows servers. Your smb.11Harrison_ch10. A value of 65 usually makes the Samba server win.

qxd 2/25/05 10:02 AM Page 160 160 Windows. Makes new files created by the user to have 644 permissions. This is actually a virtual drive that maps to the corresponding Linux users’ login directories on the Linux PDC. Allows Samba users to also write to their Samba Linux directories. 2.3 Minimum Settings for [home] Section of smb. You want to change this to 0700 so that only the login user has access to directories. The [homes] section governs how Samba handles default login directories. Samba considers all directories to be shares that can be configured with varying degrees of security. 4. you’ll see a new drive. Make your changes. Linux. usually named Z:. Choose the [homes] share section. When you log into the Samba PDC. 3. added to your PC. Click the Advanced button to see all the options.conf Value No No 0664 Description Doesn’t allow others to browse the contents of the directory. You can view the contents of the configuration file by logging into the Samba server via a command prompt and using cat /etc/samba/smb. and Samba Chapter 10 The [homes] Section Part of the process of adding a user to a Samba domain requires you to create a Linux user on the Samba PDC itself. Table 10. Your smb. Click on the SWAT shares button to proceed to where shared directories are configured. Table 10. and click on the Commit Changes button when finished.3 explains the minimum settings you need to create a functional [homes] section.conf file should resemble this when finished.11Harrison_ch10. [homes] read only = No browseable = No create mask = 0644 directory mask = 0755 . Parameter browseable read only create mask directory mask 0775 Here’s how to set the values using SWAT: 1. Makes new subdirectories created by the user to have 775 permissions. You want to change this to 0600 so that only the login user has access to files.conf to verify your changes as you do them.

conf file should look like this when you’re finished: [netlogon] path = /home/samba/netlogon guest ok = Yes [profiles] path = /home/samba/profiles read only = No create mask = 0600 directory mask = 0700 Here’s how to do it: 1. Modify the path and guest ok settings. The mask settings allow only the owner of the netlogon subdirectory to be able to modify its contents. Create a [profiles] share section. Create a [netlogon] share section. Remember to create these share directories from the command line afterwards: [root@bigboy tmp]# mkdir -p /home/samba/netlogon [root@bigboy tmp]# mkdir -p /home/samba/profile [root@bigboy tmp]# chmod -r 0755 /home/samba The [printers] Share Section Samba has special shares just for printers. The [profiles] share section stores such things as favorites and desktop icons. 2. Modify the path. 3. 5. and these are configured in the [printers] section of SWAT. and read only settings. Click the Shares button. Click on the Commit Changes button. 7. Click on the Commit Changes button. The [profiles] share section stores settings related to the look and feel of Windows so that the user has the same settings no matter which Windows PC is logged into. Your smb. The default smb. There is also a share under [printers] called printers that governs common printer settings.11Harrison_ch10. 4. 6.qxd 2/25/05 10:02 AM Page 161 Creating a Starter Configuration 161 The [netlogon] and [profiles] Share Sections The [netlogon] share section contains scripts that the Windows clients may use when they log into the domain. Print shares always have the printable parameter set to yes. mask.conf [printers] share section looks like this: .

” SAMBA PASSWORDS You should be aware that your Linux password and Samba passwords are stored in two different locations. Fortunately. as you will see throughout the chapter. the root user is the Samba administrator. “Sharing Resources Using Samba.qxd 2/25/05 10:02 AM Page 162 162 Windows. which are stored in the /etc/shadow file.conf file you saved at the beginning of this exercise has many varied examples that you may use and apply to your particular environment. Samba passwords are stored in the /etc/samba/smbpasswd file.11Harrison_ch10. Samba domain logins use the smbpasswd password. and SWAT requires you to use the Linux root password. Samba passwords are stored in the /etc/samba/smbpasswd file. . and Samba Chapter 10 [printers] comment = All Printers path = /var/spool/samba printable = Yes browseable = No Shares for Specific Groups of Users The default Samba Version 3 smb. This difference is important. or SSH require the Linux passwd command. You can find the steps for creating a simple shared directory for home users in Chapter 11. This provides the Samba administer the flexibility of allowing only some of the Linux users to have Samba accounts. Home Environment By default. HOW TO CREATE A SAMBA PDC ADMINISTRATOR USER To do both SWAT and user administration with Samba. you can add workstations to the Windows domain by creating a Samba-specific root password using the smbpasswd command. TELNET. you’ll need to create administrator accounts on the Samba PDC Linux server. and you can change them with the smbpasswd command. Linux. Use the passwd command to change Linux passwords. [root@bigboy tmp]# /usr/bin/smbpasswd -a root password Note Remember that regular Linux logins via the console.

Samba can create these Machine Trusts either manually or automatically. 4. Next. Manual Creation of Machine Trust Accounts (NT Only) The commands in the example create a special Linux group for Samba clients and then add a special machine user that’s a member of the group. Pay careful attention to the $ at the end. such as sysadmin. Create individual Linux users that are part of this group. you might want more than one person to administer Samba. via telnet or ssh.) HOW TO ADD WORKSTATIONS TO YOUR SAMBA DOMAIN Adding workstations to a Samba domain is a two-step process involving the creation of workstation trust accounts on the Samba server and then logging into each workstation to add them to the domain. Use the smbpasswd command to create Samba passwords for domain logins for this group. Create Samba Trust Accounts for Each Workstation PDCs accept user logins only from trusted PCs that have been placed in its PC client database. the commands disable the password for this user and add the machine added to the smbpasswd file to help keep track of which devices are members of the domain. Here are the steps to follow: 1. with the groupadd command. For security reasons this password may be different from the Linux password used to log into the Linux system from the console. and replace machine_name with the name of the Windows client machine. In summary. giving each an individual username. [root@bigboy tmp]# groupadd samba-clients [root@bigboy tmp]# /usr/sbin/useradd -g samba-clients \ . 2.11Harrison_ch10. Create a Linux user group.qxd 2/25/05 10:02 AM Page 163 How to Add Workstations to Your Samba Domain 163 Corporate Environment In a corporate environment.conf file so that the sysadmin group is listed in the [global] parameter settings: domain admin group = @sysadmin admin users = @sysadmin printer admin = @sysadmin 3. Use SWAT to update your smb. a machine trust account needs to have corresponding entries in the /etc/passwd and /etc/smbpasswd files. (Remember that Linux passwords are changed with the passwd command.

You can set this up by editing /etc/samba/smb. Linux.. and each version has its own procedure for joining a domain. [root@bigboy tmp]# smbpasswd -a -m machine_name Samba version 3 adds it automatically. don’t require machine trust accounts.remainder of parameters. The easiest way is to use SWAT in the Global menu to modify the add machine script parameter.> add machine script = /usr/sbin/useradd -d /dev/null -g samba-clients s /bin/false -M %u When you’ve completed the modifications. [global] # <. Make Your PC Clients Aware of Your Samba PDC There are many types of Windows installed on PCs. Here’s what you need to do: 1. .conf to automatically add the required users. you need to add the client to the smbpasswd file also. you’ll need to create the samba-clients Linux group that will help identify all the domain’s Windows clients listed in the /etc/passwd file. Windows 95/98/ME and Windows XP Home Windows 9x machines do not implement full domain membership and. [root@bigboy tmp]# groupadd samba-clients In Samba version 2.11Harrison_ch10.. Dynamic Creation of Machine Trust Accounts Although you can use the manual method. and Samba Chapter 10 -d /dev/null -s /bin/false machine_name$ [root@bigboy tmp]# passwd -l machine_name$ [root@bigboy tmp]# smbpasswd -a -m machine_name This is the only way to configure machine trusts using Windows NT. 2. therefore.qxd 2/25/05 10:02 AM Page 164 164 Windows. The next sections show you how to add the most popular versions of Windows clients to your domain. which is known as making a machine account on the fly.. Navigate to the Network section of the Control Panel (Start>Settings>Control Panel>Network). the recommended way of creating machine trust accounts is simply to allow the Samba server to create them as needed when the Windows clients join the domain.. Select the Configuration tab.

Click all the OK buttons. In this case. Check Log onto Windows NT Domain. Press the Windows and Break keys simultaneously to access the System Properties dialog.qxd 2/25/05 10:02 AM Page 165 How to Add Workstations to Your Samba Domain 165 3. 5. . You will be prompted for a user account and password with rights to join a machine to the domain. do not check the box Create a Computer Account in the Domain. 6. You can now log in using any account in the /etc/smbpasswd file with your domain as the domain name. Now. Select the Identification tab. 6. you should get a “Welcome to <DOMAIN>” message confirming that you’ve been added. 2. 5. Reboot. Click the Change button. Click OK. then go through these steps: 1. and then click OK.11Harrison_ch10. create a dynamic Samba machine trust account. Enter your domain name and computer name. 4. You can now log in using any account in the /etc/smbpasswd file with your domain as the domain name. Highlight Client for Microsoft Networks. you must first create a manual Samba machine trust account as explained earlier. simply reboot and you’re done! Windows NT For Windows NT. the user would be root with the corresponding smbpasswd password. Click the Properties button. 3. Click on the Network Identification or Computer Name tab on the top. Windows 200x and Windows XP Professional For the 200x and XP Professional varieties of Windows. 6. Enter the domain name and computer name. You should get a “Welcome to <DOMAIN>” message as confirmation that you’ve been added. Click the Properties button. 4. Navigate to the Network section of the Control Panel (Start>Settings>Control Panel>Network). 5. then follow these steps: 1. 2. Now. Reboot. In this home environment scenario. and enter the domain name. Click on the Member of Domain button. the existing machine trust account joins the machine to the domain. 7. 3. Enter the information for your Samba administrator. 4.

11Harrison_ch10.qxd

2/25/05

10:02 AM

Page 166

166

Windows, Linux, and Samba

Chapter 10

With Samba version 2, however, you may also have to make a few changes to your system’s registry using the regedit command:
[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Netlogon\ Parameters] “requiresignorseal”=dword:00000000 “signsecurechannel”=dword:00000000

After making these changes, reboot before continuing.

HOW TO ADD USERS TO YOUR SAMBA DOMAIN
Adding users to a domain has three broad phases. The first is adding a Linux user on the Samba server, the second is creating a Samba smbpasswd that maps to the new Linux user, and the third is to map a Windows drive letter to the user’s Linux home directory. Take a closer look. Adding the Users in Linux First, go through the process of adding users in Linux just as you would normally. Passwords won’t be necessary unless you want the users to log into the Samba server via TELNET or SSH. To create the user, use the command
[root@bigboy tmp]# useradd -g 100 peter

Giving them a Linux password is necessary only if the user needs to log into the Samba server directly. If the user does, use this method:
[root@bigboy tmp]# passwd peter Changing password for user peter. New password: Retype new password: passwd: all authentication tokens updated successfully. [root@bigboy tmp]#

Mapping the Linux Users to an smbpassword Next, you need to create Samba domain login passwords for the user:
[root@bigboy tmp]# /usr/bin/smbpasswd -a username password

11Harrison_ch10.qxd

2/25/05

10:02 AM

Page 167

How to Add Users to Your Samba Domain

167

The -a switch adds the user to the /etc/smbpasswd file. Use a generic password then have users change it immediately from their workstations in the usual way. Remember the smbpasswd sets the Windows domain login password for a user, which is different from the Linux login password to log into the Samba box. Mapping a Private Windows Drive Share By default, Samba automatically gives each user logged into the domain an H: drive that maps to the /home/username directory on the Linux box.
Mapping Using My Computer

If the auto-mapping doesn’t work, then try: 1. Let the user log into the domain. 2. Right-click on the My Computer icon on the desktop. 3. Click on Map Network Drive. 4. Select a drive letter. 5. Browse to the HOMENET domain, then the Samba server, then the user’s home directory. 6. Click on the check box Reconnect at Logon, to make the change permanent. If you’re more a typist than a clicker, try the next method.
Mapping from the Command Line

If you find the My Computer method too time consuming for dozens of users or if the PC doesn’t have the feature available, then you can use the commandline method and possibly make it into a script: 1. Create a master logon batch file for all users.
[root@bigboy tmp]# vi /home/samba/netlogon/login.bat

2. Add the following lines to mount the user’s share as drive P: (for private):
REM Drive Mapping Script net use P: \\bigboy\

3. Make the file world readable using:
[root@bigboy tmp]# chmod 644 /home/samba/netlogon/login.bat

4. Linux and Windows format text files differ slightly. As the file resides on a Linux box, but will be interpreted by a Windows machine, you’ll have to convert the file to the Windows format. Use the unix2dos command:
[root@bigboy tmp]# unix2dos /home/samba/netlogon/login.bat unix2dos: converting file /home/samba/netlogon/login.bat to DOS format ... [root@bigboy tmp]#

11Harrison_ch10.qxd

2/25/05

10:02 AM

Page 168

168

Windows, Linux, and Samba

Chapter 10

5. The final step is to edit your smb.conf file’s [global] section to have a valid entry for the logon script parameter. This can be done using SWAT via the Globals menu.
[global] logon script = login.bat

Now your users will have additional disk space available on a Windows P: drive whenever they login.

DOMAIN GROUPS

AND

SAMBA

Samba supports domain groups that allow users who are members of the group to have administrator rights on each PC in the domain. This enables them to add software and configure network settings, among other tasks. In Windows, domain groups also have the ability to join machines to the domain; however, Samba does not support this currently. The domain admin group parameter specifies users who will have domain administrator rights. The argument is a space-separated list of user names or group names (group names must have an @ sign prefixed). For example:
domain admin group = USER1 USER2 @GROUP

HOW TO DELETE USERS

FROM YOUR

SAMBA DOMAIN

Deleting users from your Samba domain is a two-stage process in which you have to remove the user from the Linux server and also remove the user’s corresponding smbpasswd entry: 1. Delete the users using the smbpasswd with the -x switch:
[root@bigboy tmp]# smbpasswd -x john Deleted user john. [root@bigboy root]#

2. Delete the Linux user by following the normal deletion process. For example, to delete the user john and all john’s files from the Linux server, use:
[root@bigboy tmp]# userdel -r john

Sometimes you may not want to delete the user’s files so that they can be accessed by other users at some other time. In this case, you can just deactivate the user’s account using the passwd -l username command.

11Harrison_ch10.qxd

2/25/05

10:02 AM

Page 169

Conclusion

169

HOW TO MODIFY SAMBA PASSWORDS
You can set your Samba server to allow users to make changes in their domain passwords and have these changes mirrored automatically in their Linux login passwords. Table 10.4 explains the [global] smb.conf parameters that you need to change.
Table 10.4 Parameter
unix passwd sync passwd program passwd chat

The smb.conf Settings for Enabling Online Password Changes Value Yes Use the SWAT defaults Use the SWAT defaults Description Enables Samba/Linux password synchronization. Lists the location of the Linux password file, which is usually /bin/passwd. A short script to change the Linux password using the Samba password.

CONCLUSION
By now you should have a fairly good understanding of adding users and PCs to a Samba domain. The next step is to take better advantage of Samba’s file sharing features, which is Chapter 11’s task.

11Harrison_ch10.qxd

2/25/05

10:02 AM

Page 170

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 171

C H A P T E R

11

Sharing Resources Using Samba

IN THIS CHAPTER
☞ ☞ ☞ ☞ ☞ Adding a Printer to a Samba PDC Creating Group Shares in Samba Sharing Windows Drives Using a Linux Samba Client Automating Mounting with Linux Samba Clients Conclusion

Now that you have Samba up and running, you may want to allow users to share such resources as floppy drives, directories, and printers via the Samba server. This chapter tells you how to do it all.

ADDING

A

PRINTER TO A SAMBA PDC

Sharing printers amongst all your PCs is one of the advantages of creating a home network. Here’s how to connect your printer directly to your PDC. Not only does this method make your printer available to all your Windows workstations, it also makes your Samba PDC a print server! The only potential snag is that you need the Windows printer driver loaded on every client machine. This may be okay for a small home network but impractical for a huge corporate network. Adding the Printer to Linux By far, the easiest way to add a printer in Linux is to use one of the many menu-based printer utilities available. For the example, I’ll use systemconfig-printer (Figure 11.1), which is easy to find and can be accessed from the command line:
[root@bigboy tmp]# system-config-printer

171

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 172

172

Sharing Resources Using Samba

Chapter 11

Figure 11.1

Printer configuration screen.

Note Sometimes the graphics in system-config-printer don’t work as expected due to your environment variables. You can temporarily set the required variable to the correct output with the command
[root@bigboy tmp]# env LANG=C system-config-printer

Assuming your printer is locally attached to the parallel port, here are the steps to use: 1. Using the Tab key, move to the New button and press the Enter key. 2. When the Create a New Queue menu appears, give the printer a name that’s easy to remember, select Local Printer Device, and move to the “next” button before pressing Enter. 3. From the resulting Setting Up Local Printer Device menu, select /dev/lp0, assuming the printer is on the parallel port as opposed to the USB port. Click the Next button to go to the Queue Driver menu. 4. Scroll to your manufacturer’s entry, press Enter. Scroll further to your model, and press Enter again. You’ll now get a choice of drivers, select the

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 173

Adding a Printer to a Samba PDC

173

default device, which is marked with an asterisk. Click Next to go to the Create a New Queue confirmation menu. 5. Move to the Finish button, and press Enter. There will be a slight delay. At this stage, it’s wise to do a test print to make sure all is okay. Make Samba Aware of the Printer The easiest way to let Samba know the printer is available is via the Samba SWAT Web interface. Once you are in SWAT: 1. Click the Printers button. 2. Find your printer in the pull-down menu. If the printer name has an asterisk beside it, it has been auto-configured by Samba. It might not be visible on your network, however, if Samba hasn’t been restarted since creating the printer. If this is the case, restart Samba. You can now skip ahead to the “Configure the Printer Driver on the Workstations” section. 3. If Samba did not auto-configure your printer, you need to edit or create it yourself. Click on the Commit Changes button to create an updated /etc/samba/smb.conf file. 4. Click on the Status tab at the top of the screen, and restart smbd and nmbd to restart Samba. Your printer will now be available for use by all Windows workstations. Configure the Printer Driver on the Workstations With the printer ready to go on the Linux side, you now need to prepare things in Windows. 1. Download the appropriate Windows printer driver from the manufacturer and install it. 2. Go to the Add Printer menu. Click the Next button. 3. Select the Network Printer button to access the Local or Network Printer menu. Click the Next button, again. 4. You now should be on the Locate Your Printer menu. Don’t enter a name, instead click Next so you can browse for your printer. 5. From the Browse for Printer menu, double-click on the name of your Linux Samba server. You should see the new printer. Click on the printer name, then click Next. 6. You may get the message “The server on which the printer resides does not have the correct printer driver installed. If you want to install the

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 174

174

Sharing Resources Using Samba

Chapter 11

driver on your local computer, click OK.” Fortunately, you pre-installed the driver. Click the OK button. 7. When the Add Printer Wizard appears, select the manufacturer of your printer, select the printer model, and then click OK. 8. The Add Printer Wizard will ask you whether you want to use this new printer as the default printer. Select Yes or No depending on your preference. Click the Next button 9. From the resulting Completing the Add Printer Wizard menu, click the Finish button. The new printer should now show up on the Windows Printers menu in the Control Panel. Send a test print, to be sure all is well.

CREATING GROUP SHARES

IN

SAMBA

On occasion, subgroups of a family need a share that is fully accessible by all members of the group. For example, parents working in a home office environment may need a place where they can share, distribute, or collaboratively work on documents. Here’s how it’s done. Create the Directory and User Group As with any group activity, the first step is to get organized. 1. Create a new Linux group named parents:
[root@bigboy tmp]# /usr/sbin/groupadd parents

2. Create a new directory for the group’s files. If one user is designated as the leader, you might want to change the chown statement to make that person owner:
[root@bigboy tmp]# mkdir /home/parent-files [root@bigboy tmp]# chgrp parents /home/parent-files [root@bigboy tmp]# chmod 0770 /home/parent-files

3. Add the group members to the new group. For instance, the command to add a user named father to the group is
[root@bigboy tmp]# /usr/sbin/usermod -G parents father

All your members are in the group; now they need to share. Configure the Share in SWAT Next, you need to create the share in Samba using SWAT. 1. Click on the Shares button, then enter the name of the share you want to create, such as only-parents.

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 175

Sharing Windows Drives Using a Linux Samba Client

175

2. Click on the Create Share button. Make sure the path maps to /home/ parent-files, and make the valid users be @parents, where parents is the name of the Linux user group. 3. Click on the Commit Changes button to create a new /etc/samba/smb.conf file. 4. Click on the Status tab at the top of the screen and restart smbd and nmbd to restart Samba. Your /etc/samba/smb.conf file should have an entry like this at the end
# Parents Shared Area [only-parents] path = /home/parent-files valid users = @parents

If it does, all is well and you can move on. If not, double check your work in the last steps. Map the Directory Using My Computer Finally, let the user log into the domain from a remote PC: 1. 2. 3. 4. Right click on the My Computer icon on the desktop. Click on Map Network Drive. Select a drive letter. Browse to the HOMENET domain, then the Samba server, then the share named only-parents. 5. Click on the check box Reconnect at Logon, to make the change permanent. Now the files located in the Linux /home/parent-files directory will be accessible to the parents only and your job is complete!

SHARING WINDOWS DRIVES USING

A

LINUX SAMBA CLIENT

Up to this point I have focused on your Linux server being a Samba server, but it can also mimic a Windows client using Samba’s client software. For example, you can also access a CD-ROM, DVD, Zip, floppy, or hard drive installed on a Windows machine from your Linux box. In this section I’ll show you how to share a CD-ROM drive. Windows Setup The Windows client box should be set up first as a member of a Samba domain or workgroup. The next step is to make the CD-ROM drive shared. The steps you use depend on which version of Windows you have.

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 176

176

Sharing Resources Using Samba

Chapter 11

For Windows 98/ME: 1. 2. 3. 4. Double-click My Computer. Right-click on the CD-ROM drive, and choose Sharing. Set the Share Name as cdrom with the appropriate access control. Restart Windows. For Windows 2000: 1. 2. 3. 4. Double-click My Computer. Right-click on the CD-ROM drive, and choose Sharing. Set the Share Name as cdrom with the appropriate access control. Log out, and log in again as normal using your current login. Finally, for Windows XP: 1. 2. 3. 4. Double-click My Computer. Right-click on the CD-ROM drive, and choose Sharing and Properties. Set the Share Name as cdrom, and set the appropriate access control. Log out and log in again as normal using your current login.

After you have completed this task, you’ll have to go to the next step of testing your configuration. Test Your Windows Client Configuration Use the smbclient command to test your share. You should substitute the name of your Windows client PC for “WinClient,” and in place of “username” provide a valid workgroup/domain username that normally has access to the Windows client. You should get output like this when using the username’s corresponding password:
[root@bigboy tmp]# smbclient -L WinClient -U username added interface ip=192.168.1.100 bcast=192.168.1.255 nmask=255.255.255.0 added interface ip=127.0.0.1 bcast=127.255.255.255 nmask=255.0.0.0 Got a positive name query response from 192.168.1.253 ( 192.168.1.253 ) Password: Domain=[HOMENET] OS=[Windows 5.1] Server=[Windows 2000 LAN Manager] Sharename Type Comment --------- ---- ------IPC$ IPC Remote IPC D$ Disk Default share print$ Disk Printer Drivers SharedDocs Disk cdrom Disk Printer2 Printer Acrobat PDFWriter

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 177

Sharing Windows Drives Using a Linux Samba Client

177

ADMIN$ Disk Remote Admin C$ Disk Default share Server Comment --------- ------Workgroup Master --------- -------

You can get the same result with
[root@bigboy tmp]# smbclient -L WinClient -U username%password

but this method is less secure as your password is echoed on the screen. Create a CD-ROM Drive Mount Point on Your Samba Server You’ll now need to create the mount point on the Linux server to mount and access the CD-ROM drive. In this case, I’ve named it /mnt/winclient-cdrom, and you’ll use the mount command to get access to this device from the Linux server.
Password Prompt Method

The Linux mount command will try to access the CD-ROM device as user username by using the username[eq] option. You will be prompted for a password:
[root@bigboy tmp]# mkdir /mnt/winclient-cdrom [root@bigboy tmp]# mount -t smbfs -o username=username \ //winclient/cdrom /mnt/winclient-cdrom

No Prompt Method

Linux won’t prompt you for a password if you embed the access password into the mount command string along with username:
[root@bigboy tmp]# mkdir /mnt/winclient-cdrom [root@bigboy tmp]# mount -t smbfs -o \ username=username,password=password \ //winclient/cdrom /mnt/cdrom

The smbmount Command Method

Some versions of Linux support the smbmount command to mount the remote drive. Incompatible versions will give errors like this:
[root@bigboy tmp]# smbmount //winclient/cdrom \ /mnt/winclient-cdrom -o username=username

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 178

178

Sharing Resources Using Samba

Chapter 11

Password: 27875: session setup failed: ERRDOS - ERRnoaccess (Access denied.) SMB connection failed

To be safe, stick with using the Linux mount command.

AUTOMATING MOUNTING WITH LINUX SAMBA CLIENTS
You can also automate the mounting of shares by placing entries in your /etc/fstab file. In the example below the home directory of user peter on server 192.168.1.100 will be mounted on the /mnt/smb mount point as a Samba filesystem (smbfs) using the login information in the file named /etc/cred.
# # File: /etc/fstab # //192.168.1.100/peter

/mnt/smb

smbfs

credentials=/etc/cred 0 0

The contents of the /etc/cred file needs to have the username and password for the account in this format
# # File: /etc/cred # username = peter password = peterspassword

Once finished, you can use mount -a to mount the new /etc/fstab entry, and the /mnt/smb directory will contain the contents of the share.
[root@smallfry tmp]# mount -a [root@smallfry tmp]# ls /mnt/smb backups profile docs data music [root@smallfry tmp]#

CONCLUSION
Both this chapter and the last have detailed the steps needed to configure a Samba network that is adequate for a small office or home. There are many steps to take, none are particularly complex, but you run the risk of not getting Samba to work if you omit any of them. For this reason, the next chapter is a dedicated troubleshooting guide to help you diagnose and recover from the most common Samba mistakes that we all tend to make.

in some cases.qxd 2/25/05 10:03 AM Page 179 C H A P T E R 12 Samba Security and Troubleshooting IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Testing the smb.conf File Samba and Firewall Software Testing Basic Client/Server Network Connectivity Testing Samba Client/Server Network Connectivity Checking the Samba Logs Samba Network Troubleshooting Basic Samba Security Conclusion Configuring Samba for your office or home can provide many advantages. you will usually pass the test successfully. as shown: [root@bigboy tmp]# testparm -s Load smb config files from /etc/samba/smb. TESTING THE SMB.conf Processing section “[homes]” Processing section “[printers]” 179 . This chapter explores the ways in which you can recover from those mistakes that you couldn’t avoid.conf file. and one small slip up can have big repercussions. By encouraging users to store files on a central file server you can simplify data backup and. the initial configuration of Samba can be tricky. Many simple steps need to be executed in the correct order. Unfortunately. If you used SWAT to edit the file. software installation and maintenance.CONF FILE Samba has a test utility called testparm that alerts you to errors in the smb.13Harrison_ch12.

13Harrison_ch12.qxd 2/25/05 10:03 AM Page 180 180 Samba Security and Troubleshooting Chapter 12 Loaded services file OK..255” # Local area network Broadcast Address iptables -A INPUT -i lo -j ACCEPT iptables -A OUTPUT -o lo -j ACCEPT iptables -A INPUT -p udp -s $NETWORK -d $SAMBA_SERVER \ -m multiport --dports 137. Linux iptables The Fedora installation process configures the iptables firewall package by default.0/24” # Local area network BROADCAST=”192. You have two options for working with it.1. There are other causes for Samba problems..168. Tip You can use argument: [root@bigboy tmp]# testparm -s filename testparm to test /etc/samba/smb.100 “ NETWORK=”192. which may be desirable on a secured network.. .445 -j DROP iptables -A OUTPUT -s $SAMBA_SERVER -d $NETWORK -m state --state \ ESTABLISHED. .RELATED -j ACCEPT . You can ensure that this is deactivated.168.445 -j ACCEPT iptables -A INPUT -p udp -s $NETWORK -d $BROADCAST --dport 137 \ -j ACCEPT iptables -A INPUT -p udp -d $SAMBA_SERVER -m multiport \ --dports 137. TCP ports 139) and TCP port 445 for SMB file sharing without NetBIOS. Here is sample script snippet: #!/bin/bash SAMBA_SERVER=”192. you can configure it to allow through such Microsoft protocols as NetBIOS (UDP ports 137 and 138. [root@bigboy tmp]# A successful test only means that Samba will load the configuration file..conf configuration a file that’s different from the default file. Two popular packages—iptables and Zone Alarm—offer solutions. Simply provide the filename as the first SAMBA AND FIREWALL SOFTWARE Firewall software installed both on your Windows PCs and on the Samba server may prevent Samba from functioning.1.138 -j ACCEPT iptables -A INPUT -p tcp -s $NETWORK -d $SAMBA_SERVER -m multiport \ --dports 139.168.138 -j DROP iptables -A INPUT -p tcp -d $SAMBA_SERVER -m multiport \ --dports 139.255. Or.

Double-click on the Network Connections icon. instead of an Internet network.255. After you get SAMBA to work.0 type entry. is considered as hostile traffic. and Zone Alarm should cease to interfere with Samba. 2. therefore.qxd 2/25/05 10:03 AM Page 181 Testing Basic Client/Server Network Connectivity 181 For more information. The easiest way around this is to configure Zone Alarm to consider your home network as a trusted network too.” Windows-based Zone Alarm The default installation of Zone Alarm assumes that your PC is directly connected to the Internet. please refer to Chapter 14. Click on the Advanced tab and then on the Windows Firewall Settings button. 4.13Harrison_ch12. Bring up the Control Panel. Follow these steps: 1.x/255. Turn off the Internet Connection Firewall by clearing its check box. The Windows XP Built-In Firewall You may also need to disable the firewall feature of Windows XP. but allow Windows file sharing traffic through this connection. TESTING BASIC CLIENT/SERVER NETWORK CONNECTIVITY You can perform several tests to ensure that the Samba server and all its workstations can do basic communication with each other.255.168.x. This can be done by clicking on the Exceptions tab of the Windows Firewall dialog box and clicking on the File and Printer Sharing check box. This means that the software will deny all inbound connections that attempt to connect with your PC. You may also leave the firewall on. keeping in mind the need to maintain a valid relationship with the Samba server. To do so. 5. you may want to experiment with the firewall software settings to optimize your security. 3. Make this network a trusted network. it will most likely have a 192. and select Properties. click on the Firewall tab and edit the settings for your home network. “Linux Firewalls Using iptables. The NetBIOS traffic that Samba uses to communicate with the PCs on the network. Right-click on your LAN connection icon. .

Ping the client using its DNS name. Press the Enter key when prompted for a password. or the presence of a firewall running on either the server or client.2-7.qxd 2/25/05 10:03 AM Page 182 182 Samba Security and Troubleshooting Chapter 12 From the Samba server: 1. Failure of this test may mean that Samba isn’t running on the server at all and may need to be started. [root@bigboy tmp]# smbclient -L bigboy Password: Anonymous login successful Domain=[HOMENET] OS=[Unix] Server=[Samba 3. As part of a thorough troubleshooting procedure: 1.0. 2. 3.13Harrison_ch12. from the Samba client: 1.0. Ping the server’s IP address and loopback address (127. If any of these fail. Make sure your Samba server can see all the shares available on the network with the smbclient -l samba_server command.0. TESTING SAMBA CLIENT/SERVER NETWORK CONNECTIVITY After configuring basic network connectivity.0. hosts deny. 3.1). Ping the server’s IP address.FC1] Sharename Type Comment -----------------IPC$ IPC IPC Service (Samba Server) ADMIN$ IPC IPC Service (Samba Server) Anonymous login successful Domain=[HOMENET] OS=[Unix] Server=[Samba 3. check your cabling. 2. your routing.0.1).FC1] Server --------SILENT Workgroup --------HOMENET OTHERNET [root@bigboy tmp]# Comment ------Samba Server Master ------BIGBOY SILENT . Check your SWAT configuration for invalid hosts allow. Ping the server using its DNS name. you need to go through a variety of tests to determine whether Samba has been configured correctly both on the server and client. Next. and invalid users entries.2-7.0. Ping the client’s IP address. Ping the client’s IP address and loopback address (127.

168.1.1. [root@bigboy tmp]# nmblookup -B 192.168.168.168.168.1.168. it usually fails if either your client or server has an incorrect subnet mask configured on their NIC cards.1.1.100 ( 192.100 ) Got a positive name query response from 192.168.0 querying * on 192.168.1.100 bcast=192.103 *<00> [root@bigboy tmp]# 4.255 Got a positive name query response from 192.1.13Harrison_ch12. check to see whether the client is running firewall software that could prevent communication.168. This should return the server’s IP address if the software is running correctly.103 ) Got a positive name query response from 192.1. This should return the client’s IP address if it is running correctly.100 *<00> [root@bigboy tmp]# 5.1.255 nmask=255. Another source of the problem could be that the Client for Microsoft Windows or File and Printer Sharing for Microsoft Networks settings on the client’s NIC haven’t been selected.255.168.103 192.100 __SAMBA__ querying __SAMBA__ on 192.100 ) 192. Use the smbclient //samba-server/tmp command to attempt a commandline login to the Samba server. This test actually sends out a broadcasted request for information.1.1.1.168.1.1. If the test fails.qxd 2/25/05 10:03 AM Page 183 Testing Samba Client/Server Network Connectivity 183 2. Use the nmblookup -B client-IP-address “*” command on the server to determine whether the client is accepting Samba queries.1.168.168. [root@bigboy tmp]# nmblookup -B 192.168.168.1. This should return a message that the login was successful.255.255.168. You also could have entered an incorrect IP address.0 added interface ip=192.100 __SAMBA__<00> [root@bigboy tmp]# 3.100 *<00> 192.1. Use the nmblookup -d 2 “*” command on the server to tell it to broadcast a query message to the network. Use the nmblookup -B samba-server-IP-address _SAMBA_ command on the server to determine whether the Samba software is running correctly.168.100 ( 192. If .100 bcast=192.103 “*” querying * on 192. When prompted for a password. You can test other accounts by adding the -U accountname option at the end of the command line.103 ( 192.1.255 nmask=255.100 192.168.168. [root@bigboy tmp]# nmblookup -d 2 ‘*’ added interface ip=192.103 *<00> 192.1. use the Linux password of the account with which you logged in.1.255.168. This should return answers from all locally connected clients and servers.

Make sure the corresponding Samba user has been created.FC1] tree connect failed: NT_STATUS_BAD_NETWORK_NAME [root@bigboy tmp]# A message that warns of an invalid or bad network name could mean that the tmp service on the Samba server isn’t correctly configured. Click on the Internet Properties (TCP/IP) menu option and then click on the Properties button. the username is peter. and enable Windows name resolution via DNS using the advanced TCP/IP settings menu on the PC. then make sure your hosts allow. Add the IP address of the Samba server to the WINS server settings. hosts deny. 2. C:\> .2-7.0. Click on the Network Connections icon in the Windows Control Panel. You may also need to add the name of the Samba server to the PC’s C:\WINDOWS\system32\drivers\etc\lmhosts file. 3. 6.conf file. Log into the Windows workstation as a Samba user. Messages related to bad passwords could mean that the user’s account doesn’t exist. Click on the Advanced button and then on the WINS tab.13Harrison_ch12. If it fails. This test attempts to log in using the username and password with which you logged into the PC. press the Enter key when prompted for a password.) Use the net view \\samba-server command to log into the Samba server from the command line and get a listing of your shares. [root@bigboy tmp]# smbclient //bigboy/TMP Password: Anonymous login successful Domain=[HOMENET] OS=[Unix] Server=[Samba 3.qxd 2/25/05 10:03 AM Page 184 184 Samba Security and Troubleshooting Chapter 12 you are doing this as user root. A “Network name not found” message usually points to an incorrect NetBIOS configuration on the client. Right-click on the network connection and select Properties. that their smbpasswd wasn’t created. You can get to this menu using the following method: 1. or that the password entered is incorrect. and invalid users parameters are set correctly in your smb. If you’re successful you should see: C:\>net view \\bigboy Shared resources at \\bigboy Samba Server Share name Type Used as Comment -----------------------------------------------peter Disk Home Directories The command completed successfully. (In the example below. 4.

If not. you should be able to see your domain under Windows’ My Network Places located in file manager or in the Start Menu.conf. Successful attempts should list the IP address of the master browser server. Try to map a drive letter to the user’s default login directory on the Samba server.1. You will need to set the encrypt passwords option in the [global] section of the smb.100 homenet<1d> [root@bigboy home]# This may fail with some Windows NT clients if the Samba server has been configured not to use encrypted passwords. [root@bigboy home]# nmblookup -M homenet querying fedora on 192.255 192. Once all this has tested positively. From the Samba server. Linux.conf file.168.conf file. it is sometimes best to make all your clients run similar versions of the Windows operating system.conf file to yes. and Samba. Log into the Windows workstation as a Samba user. This is done with the net use x: \\samba-server\share command.168.” newer versions of Windows send encrypted passwords only. Failure could also mean that the server’s smb.13Harrison_ch12. 8. Remember that doing so may make logins from Windows 95/98/ME clients fail. As stated in Chapter 10. Here we want user “peter” to have a DOS drive X: map to Peter’s Linux home directory on the Samba server. You should also be able to browse through the shares as well. . Make sure you have correctly configured the encrypt passwords option in the [global] section of smb. the test will fail and you will see: C:\> net view \\bigboy System error 5 has occurred. You can do this by setting the user=username option in the [tmp] section of the smb. C:\> 7.conf. C:\>net use x: \\bigboy\peter The command completed successfully. As you can see. Access is denied. you’ll need to make sure that the preferred master parameter is set to yes in the [global] section of smb. issue the nmblookup -M domain command to ensure that there is a master browser for the domain. “Windows.conf file hasn’t been configured to automatically use the PC user’s username as the Samba login name.qxd 2/25/05 10:03 AM Page 185 Testing Samba Client/Server Network Connectivity 185 If there is no user account.1. C:\> Make sure your password encryption is set correctly in the smb.

.0.0. If the network is located in the home.” discusses how to configure wireless NICs in Linux servers for this very reason.log and smbd. bind interfaces only = Yes hosts deny = ALL hosts allow = 192. you may want to hide your server where it is less intrusive due to its physical presence or to the noise of its power supply fan or hard drive.1.168.168.0/24 network. “Linux Wireless Networking. Chapter 13. ☞ Specifying the interfaces on which Samba will be active. ☞ Configuring Samba to deny all connections by default and then allowing specified hosts through with the hosts allow and hosts deny settings. try searching the nmbd.conf file.0/24 127. SAMBA NETWORK TROUBLESHOOTING It is always a good idea to use such network troubleshooting tools as tcpdump to do detailed troubleshooting. In this case. interfaces = eth0 lo CONCLUSION By now you should have a fully functional Samba-based network that is suitable for the small office or home.1. Always remember to include your loopback interface lo and the loopback interface’s network 127.qxd 2/25/05 10:03 AM Page 186 186 Samba Security and Troubleshooting Chapter 12 CHECKING THE SAMBA LOGS Samba stores all its log files in the /var/log/samba directory. If you find yourself having difficulties..0/8 in your configuration. BASIC SAMBA SECURITY You can restrict connections to your server on both a per-interface and a pernetwork basis in the [global] section of the smb.1. so I have included it here. especially if you’re not sure whether there is any bidirectional connectivity between the Samba server and the workstation. A wireless network in some cases would be ideal. You also can include the IP addresses of individual hosts in this list.log files for clues. Interface eth0 is on the 192. the 192.13Harrison_ch12. [global] .168.0/24 has been included as a valid network. This type of security is activated by: ☞ Setting the bind interfaces only parameter to yes. .

Wireless Tools is more convenient to use as it requires fewer configuration steps and the RPM package 187 . Before you consider making a Linux server wireless capable. This was how www. and wireless networks offer relatively easy and low cost deployment with security that’s becoming increasingly better. however. but at the time it wasn’t. make sure you purchase a NIC that is Linux compatible. and also because we all get stupid some of the time.14Harrison_ch13.qxd 2/25/05 10:04 AM Page 187 C H A P T E R 13 Linux Wireless Networking IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ Wireless Linux-Compatible NICs Common Wireless Networking Terms Networking with Linux Wireless Tools Networking with Linux-WLAN Troubleshooting Your Wireless LAN Wireless Networks in Businesses Conclusion My very first Linux web server was an ancient desktop computer that I bought from a secondhand store that advertised it as being “very obsolete.11 standard have many advantages— not just the aesthetic one I mentioned. Looking back. I made it wireless so I could take it out of my bedroom where the DSL connection entered my apartment. Spending more money than I should have on the antique.” It was cheap and it worked. I thought wireless Linux would be easy. This is the chapter about what started it all. The hardware is commonly available. I had so many headaches with it that I thought one of my very first Web pages should be about my little nightmare warning people about how to do it right.linuxhomenetworking. Wireless networks using the 802.com was born. but it was ugly and noisy—so noisy that it quickly became too loud to tolerate. I really did it for the challenge. You also need to decide on the wireless Linux package you intend to use: Linux-WLAN or Wireless Tools.

It is possible to buy cards with the same model number from the same vendor with very different circuitry. take time to become familiar with three key wireless terms: wireless access point. so do your research before proceeding. For this reason. but the newer version 2. Don’t worry. They then supply different drivers with each new card to make them work. which worked with Linux. you need a WAP on your network. The Linksys WMP11 wireless card is a good example of this confusion. and shared encryption key. WAPs are. which is usually connected to a regular external or integrated router for communication to the Internet. I will explain all this later. . Service Set ID. For Linux-WLAN. Learn them now. Wireless Access Points A wireless access point (WAP) is a device that acts as the central hub of all wireless data communications. analogous to switches in regular wired networks. WIRELESS LINUX-COMPATIBLE NICS Not all wireless NICs work with Linux. Before proceeding.org for the latest list of compatible hardware. all wireless servers communicate with one another via the WAP. but this prevents them from communicating with any other communications path. COMMON WIRELESS NETWORKING TERMS Learning the ins and outs of wireless Linux networks will be easier if we’re all speaking the same language. because you’ll see them throughout the chapter. You can find hardware compatibility lists for Wireless Tools quite easily on popular search engines. The original version of the card used the Intersil Prisim chip set. do your homework.7 (Broadcom chip set) and version 4 (InProComm chip set) do not. Wireless NIC manufacturers are notorious for changing the chip sets on their cards depending on the price of the components. Frequently Linux drivers for the new cards are unavailable. Servers can communicate with one another without a WAP if their NICs are configured in Ad Hoc mode.linux-wlan. For that.qxd 2/25/05 10:04 AM Page 188 188 Linux Wireless Networking Chapter 13 doesn’t have to be reinstalled every time you upgrade your kernel as with Linux-WLAN. therefore. In the most common operating mode (Infrastructure mode). check www. Even so.14Harrison_ch13. Always check the compatibility lists before buying your wireless hardware. the original WMP won’t work without upgrading the firmware. The packages support different NICs.

this makes it easy to eavesdrop on a neighboring network. As expected. Most wireless software packages enable you to view all the available ESSIDs within range and give you the option of selecting the corresponding wireless LAN (WLAN) to join. Unfortunately. installed by default. probably meets most of your 802. The Extended Service Set ID (ESSID) helps prevent the garbling of messages. Some software packages allow you to use a plain text key. Its main advantage is that.11a/b needs. unlike Linux-WLAN. NETWORKING WITH LINUX WIRELESS TOOLS The Linux Wireless Tools package. You can configure your NIC’s IP settings as if the NIC were a regular Ethernet device. . Encryption Keys Encryption is a method of encoding data so that only people with the secret key to unlock the code can view the original data. which in turn ignore all traffic using other identifiers. and therefore it is best to not only change your ESSID from the factory defaults.14Harrison_ch13. I’ll stick with ESSID unless the term SSID is relevant to an application. Each wireless network needs to be assigned an ESSID that doesn’t match that of any neighboring networks within its range of operation. you don’t have to reinstall it every time you upgrade your kernel. but the more secure ones treat the key like a password and prompt you to enter your chosen key twice without your actually being able to ever see the unencrypted key again. if you encrypt your wireless network’s data then you also need to use a shared encryption key on all the NICs and WAPs. but also to encrypt your wireless data whenever possible. The term ESSID is frequently interchangeably referred to as a Service Set ID (SSID) by many vendors.qxd 2/25/05 10:04 AM Page 189 Networking with Linux Wireless Tools 189 Extended Service Set ID The 802. Using iwconfig for Wireless Tools Configuration After physically installing your Linux-compatible NIC. After you use the ifup command the NIC becomes active. you need to configure your NIC’s IP and wireless settings before Wireless Tools works. Make sure everything works normally first and then encrypt later. but it will not function correctly as its wireless settings haven’t been configured yet. The desired ESSID is then set on both wireless NICs and WAPs. It is always best to encrypt your network data last to simplify troubleshooting.11a/b wireless networks typically found in a home environment share the same frequency range with one another so it is possible for your computer to hear the traffic meant for somebody else’s nearby network.

qxd 2/25/05 10:04 AM Page 190 190 Linux Wireless Networking Chapter 13 The most commonly used command in Wireless Tools is iwconfig.100 NETMASK=255.1. 1. # # Wireless configuration # TYPE=Wireless MODE=Managed ESSID=homenet These commands need only be on the main interface file. For example. if your wireless NIC is named eth0 and your managed network’s ESSID is homenet. provide the ESSID to use (in this case homenet). Configure your /etc/sysconfig/network-scripts/ifcfg-eth0 file normally as if it were a regular Ethernet NIC. The next section shows how to make these iwconfig changes permanent.14Harrison_ch13. however.168. iwconfig Permanent Wireless Tools Configuration Once your ad-hoc configuration has been completed. Add the following statements to the end to specify that the NIC is wireless. For the wireless mode.255. which you can use to configure most of the wireless parameters. and finally choose Managed (a WAP on present of the network) or Ad-hoc (no WAP) for the wireless mode. then the commands would be: iwconfig eth0 mode Managed iwconfig eth0 essid homenet Your NIC should now become fully functional. They are not needed for IP aliases. Your wireless NIC should function as if it were a regular Ethernet NIC using the ifup and ifdown commands. forgetting to do so can be problematic.0 ONBOOT=yes BOOTPROTO=static 2. . You will need to run these commands each time you use the ifup command. including the SSID and the wireless mode. you need to make the changes permanent. Managed means that there is a wireless access point (WAP) on the network and Ad-hoc signifies that there is none.255. DHCP Version ============ DEVICE=eth0 USERCTL=yes ONBOOT=yes BOOTPROTO=dhcp Fixed IP Version ================= DEVICE=eth0 IPADDR=192.

making it a desirable alternative based on the NIC card you have available. without the additional complications of encryption. You can use any one of the rows of characters to create a 40-bit key.14Harrison_ch13. which are composed of numeric values between 0 and 9 and alphabetic characters between A and F: [root@bigboy tmp]# /sbin/nwepgen ketchup 64:c1:a1:cc:db 2b:32:ed:37:16 b6:cc:9e:1b:37 d7:0e:51:3f:03 [root@bigboy tmp]# When using iwconfig to add encryption. You’ll notice that Linux-WLAN uses the term SSID instead of ESSID in its configuration files. then remember to use hexadecimal numbers. It is generally more difficult to install than Wireless Tools and has fewer troubleshooting tools. be sure that there are no colons or any other non-hexadecimal characters between the characters of the key. Encryption requires an encryption key that you can make up yourself or generate with the /sbin/nwepgen command that comes with the kernel-wlan-ng RPM (more on this coming up). This allows you to limit your troubleshooting activities to basic wireless settings. If you don’t have nwepgen. There should be ten characters in total: iwconfig eth0 key 967136deac The same rules (no colons or non-hexidecimals between the ten total characters) apply when using the /etc/sysconfig/network-scripts files to add encryption: KEY=967136deac NETWORKING WITH LINUX-WLAN Linux-WLAN is one of the original wireless LAN products developed for Linux. . but it does have wide ranging hardware support. The advantage of the nwepgen command is that you can provide an easily remembered string that it will consistently encode in to an ESSID key.qxd 2/25/05 10:04 AM Page 191 Networking with Linux-WLAN 191 Wireless Tools Encryption It is usually best to test your network in an unencrypted state before activating the additional security.

i386.us. remember that the filename usually starts with the software package name and a version number. Chapter 5. can find RPM Determine the Kernel Type Use the uname -p command. Identifying the Correct RPMs You http://prism2.1. ☞ You may get “device unknown” or “no such device” errors related to the wlan device in the /var/log/messages file if you use older unpatched versions of the Linux-WLAN software: Always use the most recent versions to make the installation smoother. ☞ Always be prepared to check your syslog /var/log/messages file for errors if things don’t work: It is a good source of information. but as a wlan device: This is good to know in order to avoid confusion when troubleshooting. ☞ Before installing the Linux-WLAN software for PCMCIA cards such as the Linksys WPC11 you will need to install the RPM packages that support PCMCIA: This step isn’t necessary for such true PCI cards as the Linksys WMP11.nc.” shows you how to set up syslog error logging to be more sensitive to errors. In Fedora Core.unixguru. When searching for the RPMs. “Troubleshooting Linux with syslog. Linux-WLAN doesn’t identify the wireless NIC as an Ethernet eth device. Chapter 6.31-13.rpm.14Harrison_ch13. ☞ Once configured. versions of the driver files at Remember to download the files for the correct kernel type.” is running an i686 version of Linux. The Linux version may not match the CPU you have installed. The Bigboy server discussed in Chapter 1.” covers how to do this in detail. the default SSID for your windows NICs may be different.raleigh. always use the uname version: .qxd 2/25/05 10:04 AM Page 192 192 Linux Wireless Networking Chapter 13 Linux-WLAN Preparation Here are some pointers you’ll need to remember prior to using the LinuxWLAN product: ☞ All devices on a wireless network must use the same Network Identifier or SSID to communicate with each other: The default SSID for Linux-WLAN is linux-wlan. OS version. If you need a refresher. and kernel version. Downloading and installing RPMs isn’t hard. “Installing RPM Software. as in kernel-pcmcia-cs-3. the package name is pcmcia-cs and in Red Hat 9 and earlier it is kernel-pcmcia-cs. It’s a good idea to decide on a common SSID and stick with it. “Why Host Your Own Site.

[root@bigboy tmp]# cat /etc/redhat-release Red Hat Linux release 9 (Shrike) [root@bigboy tmp]# [root@zero root]# cat /etc/fedora-release Fedora Core release 1 (Yarrow) [root@zero root]# Determine the Kernel Version You can use the uname -r command to figure out the kernel version.1-pre14.rpm kernel-wlan-ng-modules-fc1.20-8 [root@bigboy tmp]# Installing the RPMs After you have all this Linux information.1-pre14.4. In this case.2.4.rpm Note There are different RPMs for PCMCIA.and PCI-based NICs.14Harrison_ch13. Bigboy is running version 2.rpm kernel-wlan-ng-pcmcia-0. you need to download and install the base.2115-0. . remember that the filename usually starts with the software package name by a version number: kernel-wlan-ng-0.20-8: [root@bigboy tmp]# uname -r 2.1-pre14. server Bigboy is running Red Hat version 9. The base and modules RPMs need to be installed in all cases.1.i686.qxd 2/25/05 10:04 AM Page 193 Networking with Linux-WLAN 193 [root@bigboy tmp]# uname -p i686 [root@bigboy tmp]# Determine the OS Version hat-release One of the easiest ways to determine the OS version is to view the /etc/ redor the /etc/fedora-release file.i686. and interface packages. while Zero is running Fedora Core 1.2. You can also look at the /etc/issue file for other versions of Linux. When searching for the RPMs. module.i686.i686.rpm kernel-wlan-ng-pci-0.0.2.1-pre14. In this case.2.

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 194

194

Linux Wireless Networking

Chapter 13

Notice the sequence of installation in this sample output. Double-check your preparation steps and the RPM versions if the very last line of the installation gives a result code that is not success.
[root@bigboy tmp]# rpm -Uvh kernel-wlan-ng-0.2.1-pre14.i686.rpm Preparing... ##################################### [100%] 1:kernel-wlan-ng ##################################### [100%] [root@bigboy tmp]# rpm -Uvh kernel-wlan-ng-modules-fc1.1.2115-0.2.1pre14.i686.rpm Preparing... ##################################### [100%] 1:kernel-wlan-ng-modules-##################################### [100%] [root@bigboy tmp]# [root@bigboy tmp]# rpm -Uvh kernel-wlan-ng-pcmcia-0.2.1-pre14.i686.rpm Preparing... ##################################### [100%] 1:kernel-wlan-ng-pci ##################################### [100%] Adding prism2_pci alias to /etc/modprobe.conf file... ***NOTE*** YOU MUST CHANGE THIS IF YOU HAVE A PLX CARD!!! The default wlan0 network configuration is DHCP. Adjust accordingly. ACHTUNG! ATTENTION! WARNING! YOU MUST configure /etc/wlan/wlan.conf to define your SSID! YOU ALSO must configure /etc/wlan/wlancfg-SSID to match WAP settings! (---> replace SSID in filename with the value of your SSID) If you get an error after this point, there is either a problem with your drivers or you don’t have the hardware installed! If the former, get help! Starting WLAN Devices:message=dot11req_mibset mibattribute=dot11PrivacyInvoked=false resultcode=success message=dot11req_mibset mibattribute=dot11ExcludeUnencrypted=false resultcode=success [root@bigboy tmp]#

Note If you upgrade your Linux kernel, you’ll have to reinstall Linux-WLAN all over again. This will also create new versions of your /etc/sysconfig/networkscripts/ifcfg-wlan0, /etc/wlan/wlan.conf, and /etc/pcmcia/wlan-ng.opts files. You may have to restore these from the automatically saved versions. Linux-WLAN Post Installation Steps After the RPMs are installed, you need to configure the new NIC to be compatible with your network.
wlan0

wireless

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 195

Networking with Linux-WLAN

195

Configure the New wlan0 Interface

Edit /etc/sysconfig/network-scripts/ifcfg-wlan0 to include these new lines:
DHCP Version ============ DEVICE=wlan0 USERCTL=yes ONBOOT=yes BOOTPROTO=dhcp Fixed IP Version ================= DEVICE=wlan0 IPADDR=192.168.1.100 NETMASK=255.255.255.0 ONBOOT=yes BOOTPROTO=static

In the fixed IP version you also need to substitute your selected IP, netmask, network, and broadcast address with those above. Plus, make sure you have the correct gateway statement in your /etc/sysconfig/network file, for example. GATEWAY[eq]192.168.1.1.
Disable Your Existing Ethernet NIC

You may want to disable your existing eth0 Ethernet interface after installing the drivers. Add an ONBOOT=no entry to the /etc/sysconfig/networkscripts/ifcfg-eth0 file. This disables the interface on reboot or when /etc/init.d/network is restarted.
Select the Wireless Mode and SSID

All the configuration files are located in the /etc/wlan directory. The package allows your server to be connected to up to three wireless LANs. You specify the SSIDs (LAN IDs) for each wireless LAN in the /etc/wlan/wlan.conf file. In the example, I make the wlan0 interface join the homenet WLAN, as well as instruct the WLAN driver to scan all wireless channels for SSIDs.
# # Specify all the wlan interfaces on the server # WLAN_DEVICES=”wlan0” # # Specify whether the server should scan the network channels # for valid SSIDs # WLAN_SCAN=y # # Specify expected SSIDs and the wlan0 interface to which it should # be tied # SSID_wlan0=”homenet” ENABLE_wlan0=y

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 196

196

Linux Wireless Networking

Chapter 13

Each WLAN specified in the /etc/wlan/wlan.conf file has its own configuration file. Copy the /etc/wlan/wlancfg-DEFAULT file to a file named /etc/wlan/wlancfg-SSID (replace SSID with the actual SSID for your WAP). This line configures for the homenet SSID:
[root@bigboy wlan]# cp wlancfg-DEFAULT wlancfg-homenet

Start Linux-WLAN Start the wlan process and test for errors in the file /var/log/messages. All the result codes in the status messages should be success. You may receive the following error, however, which the WLAN RPM Web site claims is “harmless.”
Error for wireless request “Set Encode” (8B2A) : SET failed on device wlan0 ; Function not implemented. Error for wireless request “Set ESSID” (8B1A) : SET failed on device wlan0 ; Function not implemented.

With PCI cards, you can restart Linux-WLAN by restarting the WLAN daemon:
[root@bigboy tmp]# service wlan restart [root@bigboy tmp]# ifup wlan0

With PCMCIA cards, you can start Linux-WLAN by restarting the Linux PCMCIA daemon:
[root@bigboy tmp]# service pcmcia restart [root@bigboy tmp]# service network restart

Testing Linux-WLAN Now check to see if the IP address of the wlan0 interface is okay. Refer to the upcoming troubleshooting section if you cannot ping the network’s gateway.
[root@bigboy tmp]# ifconfig -a [root@bigboy tmp]# ping <gateway-address>

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 197

Networking with Linux-WLAN

197

Linux-WLAN Encryption for Security One of the flaws of wireless networking is that all the wireless clients can detect the presence of all available network SSIDs and have the option of joining any of them. With encryption, the client must have a membership encryption password that can also be represented as a series of Wireless Encryption Protocol (WEP) keys. The wlan.conf file (Red Hat 8.0 RPMs), wlan-SSID file (Red Hat 9/Fedora Core 1 RPMs), or /etc/pcmcia/wlan-ng.opts file (PCMCIA-type NICs) is also used to activate this feature. Tip I strongly recommend that you first set up your network without encryption. Only migrate to an encrypted design after you are satisfied that the unencrypted design works correctly. To invoke encryption, you have to set the dot11PrivacyInvoked parameter to true. You also must state which of the keys will be used as the default starting key via the dot11WEPDefaultKeyID parameter. You then have the option of either providing a key-generating string (simple password) or all four of the keys. In the example below, ketchup is the password used to automatically generate the keys.
#=======WEP=========================================== # [Dis/En]able WEP. Settings only matter if PrivacyInvoked is true lnxreq_hostWEPEncrypt=false # true|false lnxreq_hostWEPDecrypt=false # true|false dot11PrivacyInvoked=true dot11WEPDefaultKeyID=1 dot11ExcludeUnencrypted=true # true|false, in AP this means WEP # is required for all STAs # If PRIV_GENSTR is not empty, use PRIV_GENTSTR to generate # keys (just a convenience) PRIV_GENERATOR=/sbin/nwepgen # nwepgen, Neesus compatible PRIV_KEY128=false # keylength to generate PRIV_GENSTR=”ketchup” # or set them explicitly. Set genstr or keys, not both. dot11WEPDefaultKey0= # format: xx:xx:xx:xx:xx or dot11WEPDefaultKey1= # xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx dot11WEPDefaultKey2= # e.g. 01:20:03:40:05 or dot11WEPDefaultKey3= # 01:02:03:04:05:06:07:08:09:0a:0b:0c:0d

Not all devices on your network will use the same algorithm method to generate the encryption keys. You may find the same generator string will not create the same keys, rendering intra-network communication impossible. If this is the case, you can use the /sbin/nwepgen program to generate the keys after you provide an easy to remember key generator string. Once you have the four sets of keys, you’ll have to add them individually and in sequence to

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 198

198

Linux Wireless Networking

Chapter 13

wlan.conf, wlan-SSID or /etc/pcmcia/wlan-ng.opts file and set the PRIV_ GENSTR parameter to a null string of “” (the quotes are important). Here is how you can use nwepgen to create the keys with a generator string of ketchup. [root@bigboy tmp]# /sbin/nwepgen ketchup 64:c1:a1:cc:db 2b:32:ed:37:16 b6:cc:9e:1b:37 d7:0e:51:3f:03 [root@bigboy tmp]#

the

In this case, your wlan.conf or wlan-SSID file would look like this:
PRIV_GENSTR=”” # or set them explicitly. Set genstr or keys, not both. dot11WEPDefaultKey0= 64:c1:a1:cc:db dot11WEPDefaultKey1= 2b:32:ed:37:16 dot11WEPDefaultKey2= b6:cc:9e:1b:37 dot11WEPDefaultKey3= d7:0e:51:3f:03

Remember that all devices on your network, including all wireless NICs and WAPs, need to have the same keys and default key for this to work.
De-activating Encryption

In some cases, NICs without full Linux-WLAN compatibility freeze up after a number of hours of working with encryption. The steps to reverse encryption are: 1. Set the configuration file parameter dot11PrivacyInvoked to false. 2. Stop Linux-WLAN, and disable the wireless wlan0 interface.
[root@bigboy tmp]# service wlan stop Shutting Down WLAN Devices:message=lnxreq_ifstate ifstate=disable resultcode=success [root@bigboy tmp]# ifdown wlan0

3. The driver is still loaded in memory with the old encryption parameters, even though it is not active. Linux frequently loads device driver software, such as those that govern the operation of NICs, as modules that the kernel, or Linux master program, uses in its regular operation. Use the lsmod command to display a list of loaded modules. You’ll be most interested in the modules associated with 802.11 wireless protocols, which appear here as p80211 and prism2_pci:
[root@bigboy tmp]# lsmod Module Size ... ... Used by Not tainted

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 199

Troubleshooting Your Wireless LAN

199

prism2_pci p80211 ... ... [root@bigboy tmp]#

66672 20328

1 1

(autoclean) [prism2_pci]

Sometimes your NIC may use the prism drivers:
[root@bigboy tmp]# lsmod Module Size ... ... orinoco 45517 hermes 6721 ... ... [root@bigboy tmp]#

orinoco

chip set drivers instead of the

Used by

1 orinoco_pci 2 orinoco_pci,orinoco

4. Now that you have identified the driver modules in memory, unload them with the rmmod command:
[root@bigboy tmp]# rmmod prism2_pci [root@bigboy tmp]# rmmod p80211

5. Restart Linux-WLAN, reactivate the functional again:

wlan0

interface, and you should be

[root@bigboy tmp]# service wlan start Starting WLAN Devices:message=lnxreq_hostwep resultcode=no_value decrypt=false encrypt=false [root@bigboy tmp]# ifup wlan0

/var/log/messages

If you fail to reload the driver modules, you’ll get errors in your file and your NIC card will operate in an encrypted mode only:
Jan 2 18:11:12 bigboy kernel: failed,result=-110 Jan 2 18:11:18 bigboy kernel: timeout(1), reg=0x8021. Jan 2 18:11:18 bigboy kernel: failed. Jan 2 18:11:18 bigboy kernel: prism2sta_ifstate: hfa384x_drvr_start() hfa384x_docmd_wait: hfa384x_cmd hfa384x_drvr_start: Initialize command hfa384x_drvr_start: Failed, result=-110

TROUBLESHOOTING YOUR WIRELESS LAN
Linux wireless troubleshooting tools are quite extensive and provide a variety of useful information to help you get your network working. This section covers many important strategies that will complement the use of more conventional procedures such as scanning your /var/log/messages file.

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 200

200

Linux Wireless Networking

Chapter 13

Check the NIC Status When using WLAN methodology, the iwconfig, iwlist, and iwspy commands can provide useful information about the status of your wireless network. Take a closer look.
The iwconfig Command

In addition to using the regular ifconfig command to check the status of your NIC, you can use the iwconfig command to view the state of your wireless network, just don’t specify any parameters. Specifically, you can see such important information as the link quality, WAP MAC address, data rate, and encryption keys, which can be helpful in ensuring the parameters across your network are the same. For example:
[root@bigboy tmp]# iwconfig eth0 IEEE 802.11-DS ESSID:”homenet” Nickname:”bigboy” Mode:Managed Frequency:2.462GHz Access Point: 00:09:5B:C9:19:22 Bit Rate:11Mb/s Tx-Power=15 dBm Sensitivity:1/3 Retry min limit:8 RTS thr:off Fragment thr:off Encryption key:98D1-26D5-AC Security mode:restricted Power Management:off Link Quality:36/92 Signal level:-92 dBm Noise level:-148 dBm Rx invalid nwid:0 Rx invalid crypt:2 Rx invalid frag:0 Tx excessive retries:10 Invalid misc:0 Missed beacon:0 [root@bigboy tmp]#

The iwlist Command

The iwlist command can provide further information related to not just the NIC, but the entire network, including the number of available frequency channels, the range of possible data rates, and the signal strength. This example uses the command to verify the encryption key being used by the NIC, which can be very helpful in troubleshooting security related difficulties on your network:
[root@bigboy tmp]# iwlist key ... ... eth0 2 key sizes : 40, 104bits 4 keys available : [1]: 9671-36DE-AC (40 bits) [2]: off [3]: off [4]: off Current Transmit Key: [1] Security mode:open ... ... [root@bigboy tmp]#

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 201

Troubleshooting Your Wireless LAN

201

The iwlist command can verify the speed of the NIC being used, 11Mb/s in this case. This can be helpful in determining possible reasons for network slowness, especially as poor signal quality can result in the NIC negotiating a low bit rate with its WAP.
[root@bigboy tmp]# iwlist rate ... ... eth0 4 available bit-rates : 1Mb/s 2Mb/s 5.5Mb/s 11Mb/s Current Bit Rate:11Mb/s ... ... [root@bigboy tmp]#

For further information on the iwlist command, consult the man pages.
The iwspy Command

The iwspy command provides statistics on the quality of the link between your NIC and another wireless device on the network. It doesn’t run all the time; you have to activate iwspy on your interface first. When not activated, iwspy gives a “no statistics to collect” message:
[root@bigboy root]# iwspy eth0 eth0 No statistics to collect [root@bigboy root]#

To activate the command, specify the target IP address and the wireless NIC interface through which it can be found:
[root@bigboy tmp]# iwspy eth0 192.168.1.1

If you use the iwspy command without the IP address, it provides WLAN statistics with a typical/reference value against which it can be compared. In the example that follows the signal is considered fairly strong, with a 64/92 quality value versus a typical 36/92 value, but it could be weak by the historical values on your network. It’s good to check this from time to time for fluctuations.
[root@bigboy tmp]# iwspy eth0 eth0 Statistics collected: 00:09:5B:C9:19:22 : Quality:0 Signal level:0 Noise level:0 Link/Cell/AP : Quality:64/92 Signal level:-51 dBm Noise level:-149 dBm (updated)

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 202

202

Linux Wireless Networking

Chapter 13

Typical/Reference : Quality:36/92 level:-98 dBm [root@bigboy tmp]#

Signal level:-62 dBm

Noise

To switch off iwspy monitoring, add the off argument:
[root@bigboy root]# iwspy eth0 off

Check for Interrupt Conflicts Devices slotted into your PCI bus are generally assigned an interrupt value by the system, which the system uses to signal its need to communicate with the device. Multiple devices on the bus can have the same interrupt, but the system will access each one using a different memory address to avoid confusion. Sometimes this automatic allocation of interrupt (IRQ) values and memory locations is flawed and overlaps do occur, causing devices to fail. Before configuring your WLAN software, you should ensure that the wireless NIC card doesn’t have an interrupt that clashes with another device in your computer. Insert the card in an empty slot in your Linux box according to the instructions in its manual, reboot, and inspect your /var/log/messages file again:
[root@bigboy tmp]# tail -300 /var/log/messages

Look carefully for any signs that the card is interfering with existing card IRQs. If there is a conflict, there will usually be a warning or “IRQ also used by…” message. If that is the case, move the card to a different slot or otherwise eliminate the conflict by disabling the conflicting device if you don’t really need it. You should also inspect your /proc/interrupts file for multiple devices having the same interrupt:
[root@bigboy tmp]# cat /proc/interrupts 11: 4639 XT-PIC wlan0, eth0 [root@bigboy tmp]# cat /proc/interrupts 11: 4639 XT-PIC wlan0

(potentially bad)

(good)

Interrupt conflicts are usually more problematic with old style PC-AT buses; newer PCI-based systems generally handle conflicts better. The prior (potentially bad) /proc/interrupts example came from a functioning PCIbased Linux box. It worked because, although the interrupt was the same, the base memory addresses that Linux used to communicate with the cards were

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 203

Troubleshooting Your Wireless LAN

203

different. You can check both the interrupts and base memory of your NICs by using the ifconfig -a command:
[root@bigboy tmp]# ifconfig –a eth0 Link encap:Ethernet HWaddr 00:08:C7:10:74:A8 BROADCAST MULTICAST MTU:1500 Metric:1 RX packets:0 errors:0 dropped:0 overruns:0 frame:0 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:0 (0.0 b) TX bytes:0 (0.0 b) Interrupt:11 Base address:0x1820 wlan0 Link encap:Ethernet HWaddr 00:06:25:09:6A:B5 inet addr:192.168.1.100 Bcast:192.168.1.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:215233 errors:0 dropped:0 overruns:0 frame:0 TX packets:447594 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:39394014 (37.5 Mb) TX bytes:126738425 (120.8 Mb) Interrupt:11 Memory:c887a000-c887b000 [root@bigboy tmp]#

Kernel Log Errors When you find p80211 Kernel errors in /var/log/messages, they usually point to an incorrectly configured SSID or may also be caused by a NIC card with an outdated firmware version. For example:
Nov 13 22:24:54 bigboy kernel: p80211knetdev_hard_start_xmit: Tx attempt prior to association, frame dropped.

Can’t Ping Default Gateway If you can’t ping the default gateway, first check for kernel log errors. If there are no errors in /var/log/messages and you can’t ping your gateways or obtain an IP address, then check your /etc/sysconfig/networkscripts/ configuration files for a correct IP configuration and your routing table to make sure your routes are OK. You can also check to see if your Linux box is out of range of the WAP using the iwconfig command. Unknown Device Errors Look for “unknown device” or “no such device” errors in your log files or on your screen during installation or configuration. These may be caused by: ☞ An NIC that hasn’t been correctly inserted in the PCI slot. ☞ Incompatible hardware.

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 204

204

Linux Wireless Networking

Chapter 13

For example, you might see incompatible hardware errors in messages:

/var/log/

00:0c.0 Network controller: BROADCOM Corporation: Unknown device 4301 (rev01) Subsystem: Unknown device 1737:4301 Flags: bus master, fast devsel, latency 64, IRQ 5 Memory at f4000000 (32-bit, non-prefetchable) [size=3D8K] Capabilities: [40] Power Management version 2

Or, you might see errors on the screen:
Dec 1 01:28:14 bigboy insmod: /lib/modules/2.4.18-14/net/prism2_pci.o: init_module: No such device Dec 1 01:28:14 bigboy insmod: Hint: insmod errors can be caused by incorrect module parameters, including invalid IO or IRQ parameters. You may find more information in syslog or the output from dmesg Dec 1 01:28:14 bigboy insmod: /lib/modules/2.4.18-14/net/prism2_pci.o: insmod wlan0 failed

A Common Problem with Linux-WLAN and Fedora Core 1 In older versions of Fedora Core 1, the operating system will auto-detect Linux-WLAN-compatible NIC cards and enter a line similar to:
alias eth2 orinoco_pci

in the /etc/modprobe.conf file. In other words, it detects them as an Ethernet eth device instead of a WLAN wlan device. This seems to conflict with the WLAN RPMs, and you’ll get errors like this when starting Linux-WLAN:
Starting WLAN Devices: /etc/init.d/wlan: line 119: Error: Device wlan0 does not seem to be present.: command not found /etc/init.d/wlan: line 120: Make sure you’ve inserted the appropriate: command not found /etc/init.d/wlan: line 121: modules or that your modules.conf file contains: command not found /etc/init.d/wlan: line 122: the appropriate aliase(s).: command not found

You can fix the problem with the proper steps. This example refers to a compatible Orinoco chipset card: 1. Remove the orinoco_pci line from the remove the entry for device wlan0.
/etc/modprobe.conf

file. Do not

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 205

Wireless Networks In Businesses

205

2. Edit your /etc/sysconfig/hwconf file, search for orinoco_pci, and remove the orinoco_pci section that refers to your wireless card. (Each section starts and ends with a single - on a new line.) 3. Reboot. 4. The Linux boot process always runs kudzu, the program that detects new hardware. Kudzu detects the wireless card and asks whether you want to configure it. Choose ignore. This will reinsert the wireless card in the /etc/sysconfig/hwconf file, but not in the /etc/modprobe.conf file. 5. Your NIC should start to function as expected as device wlan0 when you use the ifconfig -a command. Configure the IP address, and activate the NIC as shown earlier in this chapter. The procedure removes all reference to the Orinoco driver in the Linux configuration files and then forces kudzu not to configure the NIC card according to the Linux defaults. The eth device will be recreated, but the ignore option provided to kudzu will prevent the Orinoco entry in the /etc/modprobe.conf from being reinserted, preventing conflict with the LinuxWLAN package’s wlan device.

WIRELESS NETWORKS IN BUSINESSES
Sometimes implementing a wireless network inside a business place becomes necessary. Visiting managers may need a quick connection in a conference room; sales people sharing cubicles my need it as the number of work spaces get exhausted. Perhaps someone is going to set one up on your network anyway—you might as well control this from the beginning. Apart from people who download infected software and e-mail attachments, mobile employees’ notebook computers are usually viewed as a high risk source of unintentional malicious activity as there is even less control over what these employees do than those with fixed workstations. With this in mind, it is usually best to isolate this type of wireless network completely from your internal, trusted, and wired one. Some types of network architectures make the wireless router only have access to the Internet, and no where else, via its own dedicated DSL line. The wireless users then have to use some form of a VPN client to gain access to the office servers just as if they were doing so from home. To reduce the risk of the network being hijacked, be sure to encrypt the traffic and use a proxy server running such software as Squid (see Chapter 32, “Controlling Web Access with Squid”) to limit Internet access to authorized users via some form of pop-up username and password authentication. With this sort of architecture, if the wireless network gets hijacked, your office systems should remain relatively safe. Many WAPs have the option of not advertising their ESSIDs, which prevents users from browsing around to select the nearest available WLAN. Activation of this feature can be inconvenient to users as wireless clients will

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 206

206

Linux Wireless Networking

Chapter 13

need to know the predefined ESSID to gain LAN access, but more importantly, reduces the risk of an outsider connecting to your wireless LAN by roaming the airwaves for an available WAP. Your WAP may also have the additional security feature of only allowing connections from a list of predefined MAC addresses of all the wireless NICs in your company or department. The more expensive ones have the ability to refer to a database server for this information. There are many other types of wireless methodologies. Please investigate a variety of options before coming to a final conclusion.

CONCLUSION
With the knowledge gained in the chapters in Part 1 of the book you will be able to configure a Linux file and DHCP server on small networks with relative ease. Part 2 will explore the possibility of making your server also become the core of your self-managed dedicated Web site.

15Harrison_Pt2.qxd

2/25/05

10:04 AM

Page 207

P A R T

II

The Linux Web Site Project

207

15Harrison_Pt2.qxd

2/25/05

10:04 AM

Page 208

16Harrison_ch14.qxd

2/25/05

10:06 AM

Page 209

C H A P T E R

14

Linux Firewalls Using iptables

IN THIS CHAPTER
☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ What Is iptables? Download and Install the iptables Package How to Start iptables Determining the Status of iptables Packet Processing in iptables Targets and Jumps Important iptables Command Switch Operations Using User-Defined Chains Saving your iptables Scripts Fedora’s iptables Rule Generator Recovering from a Lost Script Loading Kernel Modules Needed by iptables Sample iptables Scripts Troubleshooting iptables Conclusion

Network security is a primary consideration in any decision to host a Web site as the threats are becoming more widespread and persistent every day. One means of providing additional protection is to invest in a firewall. Though prices are always falling, in some cases you may be able to create a comparable unit using the Linux iptables package on an existing server for little or no additional expenditure. This chapter shows how to convert a Linux server into: ☞ A firewall while simultaneously being your home Web site’s mail, Web, and DNS server
209

16Harrison_ch14.qxd

2/25/05

2:37 PM

Page 210

210

Linux Firewalls Using iptables

Chapter 14

☞ A router that will use NAT and port forwarding to both protect your home network and have another Web server on your home network while sharing the public IP address of your firewall Creating an iptables firewall script requires many steps, but with the aid of the sample tutorials, you should be able to complete a configuration relatively quickly.

WHAT IS

IPTABLES?

Originally, the most popular firewall/NAT package running on Linux was ipchains, but it had a number of shortcomings. To rectify this, the Netfilter organization decided to create a new product called iptables, giving it such improvements as: ☞ Better integration with the Linux kernel with the capability of loading iptables-specific kernel modules designed for improved speed and reliability. ☞ Stateful packet inspection. This means that the firewall keeps track of each connection passing through it and in certain cases will view the contents of data flows in an attempt to anticipate the next action of certain protocols. This is an important feature in the support of active FTP and DNS, as well as many other network services. ☞ Filtering packets based on a MAC address and the values of the flags in the TCP header. This is helpful in preventing attacks using malformed packets and in restricting access from locally attached servers to other networks in spite of their IP addresses. ☞ System logging that provides the option of adjusting the level of detail of the reporting. ☞ Better network address translation. ☞ Support for transparent integration with such Web proxy programs as Squid. ☞ A rate limiting feature that helps iptables block some types of denial of service (DoS) attacks. Considered a faster and more secure alternative to ipchains, iptables has become the default firewall package installed under Red Hat and Fedora Linux.

91. “Installing RPM Software. remember that the filename usually starts with the software package name by a version number. and restart iptables after booting by using the commands: [root@bigboy tmp]# service iptables start [root@bigboy tmp]# service iptables stop [root@bigboy tmp]# service iptables restart To get iptables configured to start at boot. The first is the mangle table.i386.16Harrison_ch14. Each of these queues is dedicated to a particular type of packet activity and is controlled by an associated packet transformation/ filtering chain. . (See Chapter 6. For example: [root@bigboy tmp]# service iptables status Firewall is stopped. HOW TO START IPTABLES You can start. as in iptables-1. This is hardly used in a home or SOHO environment. There are three tables in total. stop.rpm.0. Fedora Core will give a simple status message.qxd 2/25/05 10:06 AM Page 211 Packet Processing in iptables 211 DOWNLOAD AND INSTALL THE IPTABLES PACKAGE Before you begin.” if you need a refresher. which is responsible for the alteration of quality of service bits in the TCP header. use the chkconfig command: [root@bigboy tmp]# chkconfig iptables on DETERMINING THE STATUS OF IPTABLES You can determine whether iptables is running or not via the service iptables status command. you need to make sure that the iptables software RPM is installed. [root@bigboy tmp]# PACKET PROCESSING IN IPTABLES All packets inspected by iptables pass through a sequence of built-in tables (queues) for processing.) When searching for the RPMs.2.

Used with NAT of the source IP address using either one to one or many to one NAT. Filters packets originating from the firewall. ☞ POSTROUTING chain: NATs packets when the source address of the packet needs to be changed. It has three built-in chains in which you can place your firewall policy rules: ☞ FORWARD chain: Filters packets to servers protected by the firewall. ☞ OUTPUT chain: Filters packets originating from the firewall. The third table is the nat queue. This is known as source NAT or SNAT. also known as destination NAT or DNAT. This implies that there was no need to modify the destination IP address of the packet as in prerouting.qxd 2/25/05 10:06 AM Page 212 212 Linux Firewalls Using iptables Chapter 14 The second table is the filter queue. Facilitates the transformation of the destination IP address to be compatible with the firewall’s routing table. Address translation occurs after routing.1 Processing for Packets Routed by the Firewall Queue Function Packet Transformation Chain in Queue FORWARD INPUT OUTPUT NAT Network Address Translation PREROUTING Chain Function Queue Type filter Packet filtering POSTROUTING Filters packets to servers accessible by another NIC on the firewall. . Filters packets destined to the firewall. It has two built-in chains: ☞ PREROUTING chain: NATs packets when the destination address of the packet needs to be changed.16Harrison_ch14. Table 14.1 provides more details on each queue. ☞ INPUT chain: Filters packets destined for the firewall. which is responsible for network address translation. Table 14. Used with NAT of the destination IP address. which is responsible for packet filtering. Address translation occurs before routing.

so iptables assumes that any chain that’s defined without an associated table will be a part of the filter table. It is then routed. INPUT. (Rarely used in SOHO environments. if any.) Modification of the TCP packet quality of service bits before routing occurs. if any. then it is filtered by the rules in the INPUT chain of the filter table before being processed by the intended application on the firewall. It is then inspected by the rules in the nat table’s PREROUTING chain to see whether the packet requires DNAT. OUTPUT. The packet is first examined by your rules in the mangle table’s PREROUTING chain. To help make this clearer. the packet undergoes the same sequence of steps.16Harrison_ch14. the packet undergoes SNAT before arriving at Network B.qxd 2/25/05 10:06 AM Page 213 Packet Processing in iptables 213 Queue Type Queue Function Packet Transformation Chain in Queue OUTPUT Chain Function mangle TCP header modification PREROUTING. In Figure 14.1 a TCP packet from the Internet arrives at the firewall’s interface on Network A to create a data connection. There is an exception: Most rules are related to filtering. if necessary. FORWARD Network address translation for packets generated by the firewall. If the packet is destined for the firewall itself. take a look at the way packets are handled by iptables. The filter table is therefore the default. When the destination server decides to reply. .) You need to specify the table and the chain for each firewall rule you create. then it is filtered by the rules in the FORWARD chain of the filter table and. (Rarely used in SOHO environments. POSTROUTING. At some point. If the packet is destined for a protected network. The rules in the OUTPUT chain of the nat table determine whether address translation is required and the rules in the OUTPUT chain of the filter table are then inspected before the packet is routed back to the Internet. This reply is inspected by your rules in the OUTPUT chain of the mangle table. the firewall needs to reply.

1 A diagram of the iptables packet flow. .16Harrison_ch14.qxd 2/25/05 10:06 AM Page 214 214 Linux Firewalls Using iptables Chapter 14 Packet In NetworkA Routing mangle Table PREROUTING Chain filter Table OUTPUT Chain nat Table PREROUTING Chain nat Table OUTPUT Chain Routing mangle Table OUTPUT Chain Data for the Firewall? No Yes Firewall Reply filter Table FORWARD Chain Local Processing of Data filter Table INPUT Chain nat Table POSTROUTING Chain Packet Out Network B Figure 14. It is now time to discuss the ways in which you add rules to these chains.

the packet needs to jump over to it for further processing. The packet information is sent to the syslog daemon for logging. The packet is blocked. Works like the DROP target. N/A --log-prefix “string” Tells iptables to prefix all log messages with a user defined string. rewriting the source IP address of the packet. Qualifiers include icmp-port-unreachable (default) icmp-net-unreachable icmp-host-unreachable icmp-proto-unreachable icmp-net-prohibited icmp-host-prohibited tcp-reset echo-reply DNAT Used to do destination network address translation. it is common to have two similar rules in sequence.2 lists the built-in targets that iptables uses. Most Common Options N/A DROP LOG REJECT The packet is handed over to the end application or the operating system for processing.qxd 2/25/05 10:06 AM Page 215 Targets and Jumps 215 TARGETS AND JUMPS Each firewall rule inspects each IP packet and then tries to identify it as the target of some sort of operation. --to-destination ipaddress Tells iptables what the destination IP address should be. continues . iptables stops further processing. The first logs the packet. rewriting the destination IP address of the packet. the second drops it. but also returns an error message to the host sending the packet that the packet was blocked. Table 14. Used to do source network address translation.2 Target ACCEPT Descriptions of the Most Commonly Used Targets Description iptables stops further processing. Table 14. iptables continues processing with the next rule in the table. Once a target is identified. As you can’t log and drop at the same time. --to-source <address> [-<address>][:<port>-<port>] SNAT The source IP address is user defined. Specifies the source IP address and ports to be used by SNAT. --reject-with qualifier The qualifier tells what type of reject message is returned. Frequently used to tell why the logged packet was dropped.16Harrison_ch14.

Types include icmp. Tables 14.16Harrison_ch14. Jump to the specified target chain when the packet matches the current rule. Match protocol. tcp. The possible built-in tables include: filter. The 0/0 representation of an IP address means any. such the source IP address and TCP port.1. Match destination IP address. iptables . Match input interface on which the packet enters. Match output interface on which the packet exits. Most Common Options [--to-ports <port>[-<port>]] Target MASQUERADE Specifies the range of source ports to which the original source port can be mapped.3 through 14.168.qxd 2/25/05 10:06 AM Page 216 216 Linux Firewalls Using iptables Chapter 14 Table 14. Flush. nat.7 list the most common options. Table 14. IMPORTANT IPTABLES COMMAND SWITCH OPERATIONS Each line of an iptables script not only has a jump. but they also have a number of command line options that are used to append rules to chains that match your defined packet characteristics. iptables Command Switch -t <table> -j <target> -A -F -p <protocol-type> -s <ip-address> -d <ip-address> -i <interface-name> -o <interface-name> In this command switches example: iptables -A INPUT -s 0/0 -i eth0 -d 192. There are also options that can be used to just clear a chain so you can start all over again.168. mangle. Deletes all the rules in the selected table. udp and all. then the filter table is assumed.1 -p TCP -j ACCEPT is being configured to allow the firewall to accept TCP packets coming in on interface eth0 from any IP address destined for the firewall’s IP address of 192. By default the source IP address is the same as that used by the firewall’s interface. Append rule to end of a chain. Match source IP address.1.2 continued Description Used to do source network address translation.1.3 General iptables Match Criteria Description If you don’t specify a table.

You can specify time intervals in the format .168. The source port is in the range 1024 to 65535 and the destination port is port 80 (www/http).16Harrison_ch14. Common ICMP (Ping) Match Criteria Description The most commonly used types are echo-reply and echo-request Table 14.qxd 2/25/05 10:06 AM Page 217 Important iptables Command Switch Operations 217 Table 14. can be a single value or a range in the format:starting-port: ending-port -p tcp --sport <port> -p tcp --dport <port> -p tcp --syn -p udp --sport <port> -p udp --dport <port> In this example: iptables -A FORWARD -s 0/0 -i eth0 -d 192. Consider another example: iptables -A INPUT -p icmp --icmp-type echo-request \ -m limit --limit 1/s -i eth0 -j ACCEPT The limit feature in iptables specifies the maximum average number of matches to allow per second.5 Matches used with ---icmp-type --icmp-type <type> In this example: iptables -A OUTPUT -p icmp --icmp-type echo-request -j ACCEPT iptables -A INPUT -p icmp --icmp-type echo-reply -j ACCEPT iptables is being configured to allow the firewall to send ICMP echo-requests (pings) and in turn. accept the expected ICMP echo-replies.1.58 that is reachable via interface eth1.4 Switch Common TCP and UDP Match Criteria Description TCP source port. not a new connection request UDP source port.1. can be a single value or a range in the format:start-port-number: end-port-number TCP destination port. ! --syn means. can be a single value or a range in the format: starting-port: ending-port Used to identify a new connection request.168.58 -o eth1 -p TCP \ --sport 1024:65535 --dport 80 -j ACCEPT iptables is being configured to allow the firewall to accept TCP packets for routing when they enter on interface eth0 from any IP address and are destined for an IP address of 192. Can be a single value or a range in the format: starting-port: ending-port UDP destination port.

The most frequently tested states are: ESTABLISHED: The packet is part of a connection that has seen packets in both directions.58 -o eth1 -p TCP \ --sport 1024:65535 -m multiport --dport 80. A variety of TCP/UDP ports separated by commas. When tuned correctly.16Harrison_ch14. A variety of TCP/UDP destination ports separated by commas.1. Unlike when -m isn’t used. port> -m --ports <port. Here a defense for SYN flood attacks was created by limiting the acceptance of TCP segments with the SYN bit set to no more than five per second. NEW: The packet is the start of a new connection.1. you can use abbreviations so that 3/second is In this example.443 -j ACCEPT iptables -A FORWARD -d 0/0 -o eth0 -s 192. or /day. Table 14.168. port> -m --state <state> This is an expansion on the previous example: iptables -A FORWARD -s 0/0 -i eth0 -d 192.qxd 2/25/05 10:06 AM Page 218 218 Linux Firewalls Using iptables Chapter 14 /second. iptables -A INPUT -p tcp --syn -m limit --limit 5/s -i eth0 -j ACCEPT You can expand on the limit feature of iptables to reduce your vulnerability to certain types of denial of service attacks. ICMP echo requests are restricted to no more than one per second. -m --dport <port. or the same as 3/s. INVALID: The packet couldn’t be identified. /minute. Could be due to insufficient system resources or ICMP errors that don’t match an existing data flow.168. This is a common feature of such protocols as an FTP data transfer. /hour. or an ICMP error. RELATED: The packet is starting a new secondary connection. this feature allows you to filter unusually high volumes of traffic that characterize denial of service (DOS) attacks and Internet worms. Source and destination ports are assumed to be the same and they do not have to be within a range.6 Switch -m --sport <port.58 -i eth1 -p TCP \ -m state --state ESTABLISHED -j ACCEPT . Unlike when -m isn’t used. they do not have to be within a range. port> Common Extended Match Criteria Description A variety of TCP/UDP source ports separated by commas. they do not have to be within a range.

110. For example: iptables -A INPUT -i eth0 -d 206. you can simply allow packets related to established connections using the -m state and --state ESTABLISHED options.229.qxd 2/25/05 10:06 AM Page 219 Using User-Defined Chains 219 Here iptables is being configured to allow the firewall to accept TCP packets to be routed when they enter on interface eth0 from any IP address destined for IP address of 192.168. instead of using a single.8 Chain INPUT OUTPUT fast-input-queue fast-output-queue icmp-queue-out icmp-queue-in Custom Queues Example Listing Description The regular built-in INPUT chain in iptables The regular built-in OUTPUT chain in iptables Input chain dedicated to identifying specific protocols and shunting the packets to protocol specific chains Output chain dedicated to identifying specific protocols and shunting the packets to protocol specific chains Output queue dedicated to ICMP Input queue dedicated to ICMP .16Harrison_ch14. built-in chain for all protocols.110. thereby shortening the total length of all chains the packet has to pass through. Table 14.8 summarizes the function of each.229. you can replace a long chain with a stubby main chain pointing to multiple stubby chains.1. In other words. Table 14.168. protocol-specific chain in the filter table. Instead of stating the source and destination ports.58 that is reachable via interface eth1. you can configure iptables to have user-defined chains. This feature is frequently used to help streamline the processing of packets.1.2 -j fast-output-queue iptables -A fast-input-queue -p icmp -j icmp-queue-in iptables -A fast-output-queue -p icmp -j icmp-queue-out iptables -A icmp-queue-out -p icmp --icmp-type echo-request \ -m state --state NEW -j ACCEPT iptables -A icmp-queue-in -p icmp --icmp-type echo-reply -j ACCEPT Here six queues help assist in improving processing speed.2 -j fast-input-queue iptables -A OUTPUT -o eth0 -s 206.58 are allowed to be accepted too. USING USER-DEFINED CHAINS As you may remember. For example. The return packets from 192. The source port is in the range 1024 to 65535 and the destination ports are port 80 (www/http) and 443 (https). you can use the chain to determine the protocol type for the packet and then hand off the actual final processing to a user-defined.

For these reasons.qxd 2/25/05 10:06 AM Page 220 220 Linux Firewalls Using iptables Chapter 14 SAVING YOUR IPTABLES SCRIPTS The service iptables save command permanently saves the iptables configuration in the /etc/sysconfig/iptables file. . The initialization of built-in chains is automatic and the string “iptables” is omitted from the rule statements. IPSec (ESP and AH packets). already established connections. It prompts for the level of security and then gives you the option of doing simple customizations. Like the service iptables save command. FEDORA’S IPTABLES RULE GENERATOR Fedora comes with a program called lokkit that you can use to generate a very rudimentary firewall rule set.9 on Mon Nov 8 11:00:07 2004 *filter :INPUT ACCEPT [0:0] :FORWARD ACCEPT [0:0] :OUTPUT ACCEPT [144:12748] :RH-Firewall-1-INPUT . It is a good place for beginners to start on a test system so that they can see a general rule structure.[0:0] -A INPUT -j RH-Firewall-1-INPUT -A FORWARD -j RH-Firewall-1-INPUT -A RH-Firewall-1-INPUT -i lo -j ACCEPT -A RH-Firewall-1-INPUT -p icmp -m icmp --icmp-type 255 -j ACCEPT -A RH-Firewall-1-INPUT -p esp -j ACCEPT -A RH-Firewall-1-INPUT -p ah -j ACCEPT -A RH-Firewall-1-INPUT -m state --state RELATED. lokkit saves the firewall rules in a new /etc/sysconfig/iptables file for use on the next reboot. you’re better off writing and applying a customized script and then using the service iptables save command to make the changes permanent. Here is a sample /etc/sysconfig/iptables configuration that allows ICMP.16Harrison_ch14.2.ESTABLISHED -j ACCEPT -A RH-Firewall-1-INPUT -p tcp -m state --state NEW -m tcp --dport 22 j ACCEPT -A RH-Firewall-1-INPUT -j REJECT --reject-with icmp-host-prohibited COMMIT # Completed on Mon Nov 8 11:00:07 2004 [root@bigboy tmp]# It is never a good idea to edit this script directly because it is always overwritten by the save command and it doesn’t save any comments at all. the iptablesrestore program reads the configuration and makes it the active configuration. The format of the /etc/sysconfig/iptables file is slightly different from that of the scripts shown in this chapter. which can also make it extremely difficult to follow. When the system reboots. and inbound SSH: [root@bigboy tmp]# cat /etc/sysconfig/iptables # Generated by iptables-save v1.

you can use the iptables-save and iptables-restore commands to assist you with the continued management of the server. Unlike the service iptables save command.16Harrison_ch14. If you redirect the iptables-save screen output to a file with the > symbol. In these situations.[0:0] -A INPUT -j RH-Firewall-1-INPUT -A FORWARD -j RH-Firewall-1-INPUT -A RH-Firewall-1-INPUT -i lo -j ACCEPT -A RH-Firewall-1-INPUT -p icmp -m icmp --icmp-type 255 -j ACCEPT -A RH-Firewall-1-INPUT -p esp -j ACCEPT -A RH-Firewall-1-INPUT -p ah -j ACCEPT -A RH-Firewall-1-INPUT -m state --state RELATED. This example exports the iptables-save output to a text file named firewall-config: [root@bigboy tmp]# iptables-save > firewall-config [root@bigboy tmp]# cat firewall-config # Generated by iptables-save v1. RECOVERING FROM A LOST SCRIPT Sometimes the script you created to generate iptables rules may get corrupted or lost.2. then you can edit the output and reload the updated rules when they meet your new criteria with the iptables-restore command. it’s best to write scripts that you can comment and then save them to /etc/sysconfig/ iptables.qxd 2/25/05 10:06 AM Page 221 Recovering from a Lost Script 221 Once you have become familiar with the iptables syntax. iptables-save displays the active configuration to the screen in /etc/sysconfig/iptables format. It makes them much more manageable and readable.9 on Mon Nov 8 11:00:07 2004 *filter :INPUT ACCEPT [0:0] :FORWARD ACCEPT [0:0] :OUTPUT ACCEPT [144:12748] :RH-Firewall-1-INPUT . which actually saves a permanent copy of the firewall’s active configuration in the /etc/sysconfig/iptables file. or you might inherit a new system from an administer and cannot find the original script used to protect it. you can reload it into the active firewall rule set with the iptables-restore command: [root@bigboy tmp]# iptables-restore < firewall-config .ESTABLISHED -j ACCEPT -A RH-Firewall-1-INPUT -p tcp -m state --state NEW -m tcp --dport 22 j ACCEPT -A RH-Firewall-1-INPUT -j REJECT --reject-with icmp-host-prohibited COMMIT # Completed on Mon Nov 8 11:00:07 2004 [root@bigboy tmp]# After editing the firewall-config file with the commands you need.

The ip_conntrack_ftp module needs to be added for FTP support and should always be loaded with the ip_conntrack module. which is run at the end of every reboot.qxd 2/25/05 10:06 AM Page 222 222 Linux Firewalls Using iptables Chapter 14 Finally. you should permanently save the active configuration so that it will be loaded automatically when the system reboots: [root@bigboy tmp]# service iptables save If desired. LOADING KERNEL MODULES NEEDED BY IPTABLES The iptables application requires you to load certain kernel modules to activate some of its functions. The script samples in this chapter include these statements only as a reminder to place them in the /etc/rc. which tracks TCP connection states. The ip_nat_ftp module also needs to be loaded for FTP servers behind a NAT firewall. so you’ll have to add the statements to your /etc/rc.16Harrison_ch14.local file. the iptable_nat module needs to be loaded. you can eventually convert this firewall-config file into a regular iptables script so that it becomes more easily recognizable and manageable. As most scripts probably will keep track of connection states.local file: # File: /etc/rc. the /etc/sysconfig/iptables file doesn’t support the loading of modules. Pay special attention to the logging example at the end.local # Module to track the state of connections modprobe ip_conntrack # Load the iptables active FTP module. requires ip_conntrack modprobe ip_conntrack_ftp # Load iptables NAT module when required modprobe iptable_nat # Module required for active an FTP server using NAT modprobe ip_nat_ftp SAMPLE IPTABLES SCRIPTS This section provides some sample scripts you can use to get iptables working for you. Unfortunately. . Whenever any type of NAT is required. the ip_conntrack module will be needed in any case.

” to find detailed scripts.conf to modify kernel parameters is explained in more detail in Appendix I. Note Once you feel more confident. Scripts. mail. Other snippets will help you get basic functionality. the Linux operating system has a number of built-in protection mechanisms that you should activate by modifying the system kernel parameters in the /proc filesystem via the /etc/sysctl. and ICMP packets ☞ Provide outbound passive FTP access from the firewall There are also simpler code snippets in Appendix II for Inbound and Outbound FTP connections to and from your firewall. Helps to maintain state # Also protects against IP spoofing #--------------------------------------------------------------net/ipv4/conf/all/rp_filter = 1 #--------------------------------------------------------------# Enable logging of packets with malformed IP addresses #--------------------------------------------------------------net/ipv4/conf/all/log_martians = 1 . and Configurations.” Here is a sample configuration: # File: /etc/sysctl.qxd 2/25/05 10:06 AM Page 223 Sample iptables Scripts 223 The basic initialization script snippet should also be included in all your scripts to ensure the correct initialization of your chains should you decide to restart your script after startup.conf #--------------------------------------------------------------# Disable routing triangulation. and DNS server ☞ Be the NAT router for your home network ☞ Prevent various types of attacks using corrupted TCP. you can use Appendix II.16Harrison_ch14. “Codes. UDP. It should be a good guide to get you started. “Miscellaneous Linux Topics. Respond to queries out # the same interface.conf file. For example. not another. The appendix shows you how to allow your firewall to: ☞ Be used as a Linux Web. Basic Operating System Defense You can do several things before employing your firewall script to improve the resilience of your firewall to attack. Using of /etc/sysctl.

ipv4.ipv4..conf..all.ipv4.log_martians = 1 net.ipv4. net.conf. NAT included #--------------------------------------------------------------net/ipv4/ip_forward = 1 This configuration will become active after the next reboot.qxd 2/25/05 10:06 AM Page 224 224 Linux Firewalls Using iptables Chapter 14 #--------------------------------------------------------------# Disable redirects #--------------------------------------------------------------net/ipv4/conf/all/send_redirects = 0 #--------------------------------------------------------------# Disable source routed packets #--------------------------------------------------------------net/ipv4/conf/all/accept_source_route = 0 #--------------------------------------------------------------# Disable acceptance of ICMP redirects #--------------------------------------------------------------net/ipv4/conf/all/accept_redirects = 0 #--------------------------------------------------------------# Turn on protection from Denial of Service (DOS) attacks #--------------------------------------------------------------net/ipv4/tcp_syncookies = 1 #--------------------------------------------------------------# Disable responding to ping broadcasts #--------------------------------------------------------------net/ipv4/icmp_echo_ignore_broadcasts = 1 #--------------------------------------------------------------# Enable IP routing. ..conf. but changes won’t become active in the current boot session until you run the sysctl -p command: [root@bigboy tmp]# sysctl -p .conf.icmp_echo_ignore_broadcasts = 1 [root@bigboy tmp]# .all.rp_filter = 1 net. Required if your firewall is protecting a # network.all.accept_source_route = 0 net..ipv4.16Harrison_ch14.all.all.tcp_syncookies = 1 net.conf.accept_redirects = 0 net.ipv4.ipv4.send_redirects = 0 net.

and OUTPUT chains for processing. It is not good policy. The filter table’s INPUT. the user defined # chains should be deleted. then they will not reach the INPUT. This value will not be retained in the # /etc/sysconfig/iptables file. then # The policy should be to drop it #--------------------------------------------------------------iptables iptables iptables iptables iptables --policy INPUT -j DROP --policy OUTPUT -j DROP --policy FORWARD -j DROP -t nat --policy POSTROUTING ACCEPT -t nat --policy PREROUTING ACCEPT #--------------------------------------------------------------# The loopback interface should accept all traffic . to initialize your chain and table settings with known values.local example in this # chapter. FORWARD. Included only as a reminder. and if all packets that don’t match the nat and mangle table rules are dropped. #--------------------------------------------------------------modprobe modprobe modprobe modprobe ip_conntrack ip_nat_ftp ip_conntrack_ftp iptable_nat #--------------------------------------------------------------# Initialize all the chains by removing all the rules # tied to them #--------------------------------------------------------------iptables –flush iptables -t nat –flush iptables -t mangle –flush #--------------------------------------------------------------# Now that the chains have been initialized. We’ll recreate them in the next step #--------------------------------------------------------------iptables --delete-chain iptables -t nat --delete-chain iptables -t mangle --delete-chain #--------------------------------------------------------------# If a packet doesn’t match one of the built in chains. Additional ALLOW rules should be added to the end of this script snippet. FORWARD.qxd 2/25/05 10:06 AM Page 225 Sample iptables Scripts 225 Basic iptables Initialization It is a good policy. and OUTPUT chains should drop packets by default for the best security.16Harrison_ch14. in any iptables script you write. #--------------------------------------------------------------# Load modules for FTP connection tracking and NAT . however.You may need # them later # # Note: It is best to use the /etc/rc. These tables are queried before the filter table. to make your nat and mangle tables drop packets by default.

You can take even more precautions to further protect your network.158.qxd 2/25/05 10:06 AM Page 226 226 Linux Firewalls Using iptables Chapter 14 # Necessary for X-Windows and other socket based services #--------------------------------------------------------------iptables -A INPUT -i lo -j ACCEPT iptables -A OUTPUT -o lo -j ACCEPT Advanced iptables Initialization You may also want to add some more advanced initialization steps to your script.16Harrison_ch14. This removes the need to repeat the same statements over and over again.25” # External Internet interface # Internet Interface IP address #———————————————————————————————# Initialize our user-defined chains #———————————————————————————————iptables -N valid-src iptables -N valid-dst #———————————————————————————————# Verify valid source and destination addresses for all packets #———————————————————————————————iptables iptables iptables iptables -A -A -A -A INPUT FORWARD OUTPUT FORWARD -i -i -o -o $EXTERNAL_INT $EXTERNAL_INT $EXTERNAL_INT $EXTERNAL_INT -j -j -j -j valid-src valid-src valid-dst valid-dst #=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=# # # Source and Destination Address Sanity Checks # # Drop packets from networks covered in RFC 1918 (private nets) # Drop packets from external interface IP # . The sample script snippet below outlines how to do this. The complete firewall script in Appendix II outlines most of them. which are beyond the scope of this book. # #=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#= EXTERNAL_INT=”eth0” EXTERNAL_IP=”97. The script also uses multiple user-defined chains to make the script shorter and faster as the chains can be repeatedly accessed. including checks for Internet traffic from RFC1918 private addresses. #=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#= # # Define networks: NOTE!! You may want to put these “EXTERNAL” # definitions at the top of your script.253. More complex initializations would include checks for attacks using invalid TCP flags and directed broadcasts.

0.0/8 0.0.0.0.0/4 $240.0.0/16 $224. .0.0.255.0/8 $172.255.0. thereby making the first steps in establishing a connection.0.16. Connections initiated by users logged into the Web server will be denied as outbound NEW connection packets aren’t allowed.0. #--------------------------------------------------------------# Allow outbound DNS queries from the FW and the replies too # # . This is not because it is required for the basic functionality of the firewall. The following statements will apply not only for firewalls acting as DNS clients but also for firewalls working in a caching or regular DNS server role.254.0/12 $192. which will help to keep the server up to date with the latest security patches.0/8 255.0.0.0/4 -j -j -j -j -j -j -j -j -j -j -j DROP DROP DROP DROP DROP DROP DROP DROP DROP DROP DROP Allowing DNS Access to Your Firewall You’ll almost certainly want your firewall to make DNS queries to the Internet. It isn’t necessary to specify these ports for the return leg as outbound packets for all established connections are allowed.0.0. but because of Fedora Linux’s yum RPM updater.168.0.16Harrison_ch14.0/16 $EXTERNAL_IP $224. Inbound packets destined for ports 80 and 22 are allowed. Most home networks # / websites using a single DNS server won’t require TCP statements # #--------------------------------------------------------------iptables -A OUTPUT -p udp -o eth0 --dport 53 --sport 1024:65535 \ -j ACCEPT iptables -A INPUT -p udp -i eth0 --sport 53 --dport 1024:65535 \ -j ACCEPT Allowing WWW and SSH Access to Your Firewall This sample snippet is for a firewall that doubles as a Web server that is managed remotely by its system administrator via secure shell (SSH) sessions.255 169.0/5 $127.Interface eth0 is the internet interface # # Zone transfers use TCP and not UDP.qxd 2/25/05 10:06 AM Page 227 Sample iptables Scripts 227 #=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=# iptables iptables iptables iptables iptables iptables iptables iptables iptables iptables iptables -A -A -A -A -A -A -A -A -A -A -A valid-src valid-src valid-src valid-src valid-src valid-src valid-src valid-src valid-src valid-src valid-dst -s -s -s -s -s -s -s -d -s -s -d $10.

qxd 2/25/05 2:38 PM Page 228 228 Linux Firewalls Using iptables Chapter 14 #--------------------------------------------------------------# Allow previously established connections # .RELATED -i eth0 -p tcp \ If you want all TCP traffic originating from the firewall to be accepted. #--------------------------------------------------------------# Allow port 80 (www) and 443 (https) connections from the firewall #--------------------------------------------------------------iptables -A OUTPUT -j ACCEPT -m state --state NEW \ -o eth0 -p tcp -m multiport --dport 80. as well as by Red Hat Linux servers running up2date. Note HTTPS (secure HTTP) is used for credit card transactions frequently. then remove the line: -m multiport --dport 80. FTP and HTTP are frequently used with yum. and HTTPS uses port 443.16Harrison_ch14.443 -m multiport \ --sport 1024:65535 #--------------------------------------------------------------# Allow previously established connections # . HTTP traffic uses TCP port 80.RELATED \ -j ACCEPT #--------------------------------------------------------------# Allow port 80 (www) and 22 (SSH) connections to the firewall #--------------------------------------------------------------iptables -A INPUT -p tcp -i eth0 --dport 22 --sport 1024:65535 \ -m state --state NEW -j ACCEPT iptables -A INPUT -p tcp -i eth0 --dport 80 --sport 1024:65535 \ -m state --state NEW -j ACCEPT Allowing Your Firewall to Access the Internet This iptables script enables a user on the firewall to use a Web browser to surf the Internet.443 --sport 1024:65535 .Interface eth0 is the internet interface #--------------------------------------------------------------iptables -A INPUT -j ACCEPT -m state --state ESTABLISHED.Interface eth0 is the internet interface #--------------------------------------------------------------iptables -A OUTPUT -o eth0 -m state --state ESTABLISHED.

you will have to configure iptables to allow packets to flow between the two interfaces. Masquerading also depends on the Linux operating system being configured to support routing between the Internet and private network interfaces of the firewall.168. “Introduction to Networking. eth1 is directly connected to a home network using IP addresses from the 192. An example of this can be seen in the static NAT section that follows.0 network.1.168. use the FORWARD chain of the filter table. Further rules will be needed for the interface connected to the Internet to allow only specific ports. More specifically. All traffic between this network and the firewall is simplistically assumed to be trusted and allowed. but only packets related to . This is done by enabling IP forwarding or routing by giving the file /proc/sys/net/ipv4/ip_forward the value 1 as opposed to the default disabled value of 0.0/24 -i eth1 -j ACCEPT -p all -d 192.Interface eth1 is the private network interface #--------------------------------------------------------------iptables -A INPUT iptables -A OUTPUT -j ACCEPT -p all -s 192. traffic from all devices on one or more protected networks will appear as if it originated from a single IP address on the Internet side of the firewall. Keep in mind that iptables requires the iptables_nat module to be loaded with the modprobe command for the masquerade feature to work.1. You can configure many to one NAT to an IP alias.qxd 2/25/05 10:06 AM Page 229 Sample iptables Scripts 229 Allow Your Home Network to Access the Firewall In this example.0/24 -o eth1 Masquerading (Many to One NAT) As explained in Chapter 2. types of connections.168. packets related to NEW and ESTABLISHED connections will be allowed outbound to the Internet.16Harrison_ch14.1. Note The masquerade IP address always defaults to the IP address of the firewall’s main interface. In other words. To do this. using the POSTROUTING and not the MASQUERADE statement. and possibly even remote servers to have access to your firewall and home network. The advantage of this is that you never have to specify the NAT IP address. Once masquerading has been achieved using the POSTROUTING chain of the nat table.” masquerading is another name for what many call many to one NAT. #--------------------------------------------------------------# Allow all bidirectional traffic from your firewall to the # protected network # . This makes it much easier to configure iptables NAT with DHCP.

established and related connections # Allowed inbound : Established and related connections #--------------------------------------------------------------iptables -A FORWARD -t filter -o eth0 -m state \ --state NEW. Port Forwarding Type NAT (DHCP DSL) In many cases home users may get a single DHCP public IP address from their ISPs. This value will not be retained in the # /etc/sysconfig/iptables file. #--------------------------------------------------------------modprobe iptable_nat #--------------------------------------------------------------# Enable routing by modifying the ip_forward /proc filesystem file # # Note: It is best to use the /etc/sysctl.RELATED -j ACCEPT iptables -A FORWARD -t filter -i eth0 -m state \ --state ESTABLISHED.Interface eth0 is the internet interface # . # Allowed outbound: New.Interface eth1 is the private network interface #--------------------------------------------------------------iptables -A POSTROUTING -t nat -o eth0 -s 192.16Harrison_ch14. #--------------------------------------------------------------# Load the NAT module # # Note: It is best to use the /etc/rc.168.conf example in this # chapter. then it should be used as the default gateway for all your servers on the network.RELATED -j ACCEPT Note If you configure your firewall to do masquerading. #--------------------------------------------------------------echo 1 > /proc/sys/net/ipv4/ip_forward #--------------------------------------------------------------# Allow masquerading # .qxd 2/25/05 10:06 AM Page 230 230 Linux Firewalls Using iptables Chapter 14 ESTABLISHED connections will be allowed inbound. If a Linux firewall is also your interface to the Internet and you . the packets are routed via the filter # table’s FORWARD chain.local example in this # chapter.1. This helps to protect the home network from anyone trying to initiate connections from the Internet.0/24 -d 0/0 \ -j MASQUERADE #--------------------------------------------------------------# Prior to masquerading.ESTABLISHED. Included only as a reminder. This value will not be retained in the # /etc/sysconfig/iptables file. Included only as a reminder.

#--------------------------------------------------------------modprobe iptable_nat #--------------------------------------------------------------# Get the IP address of the Internet interface eth0 (linux only) # # You’ll have to use a different expression to get the IP address # for other operating systems which have a different ifconfig output # or enter the IP address manually in the PREROUTING statement # # This is best when your firewall gets its IP address using DHCP.168. This value will not be retained in the # /etc/sysconfig/iptables file. then you will have to use port forwarding. the remote server’s IP address.qxd 2/25/05 10:06 AM Page 231 Sample iptables Scripts 231 want to host a Web site on one of the NAT protected home servers.*://’`” #--------------------------------------------------------------# Enable routing by modifying the ip_forward /proc filesystem file # # Note: It is best to use the /etc/sysctl.local example in this # chapter. and the source/destination port of the traffic can be used to uniquely identify a traffic flow. Included only as a reminder.1. All traffic that matches a particular combination of these factors may then be forwarded to a single server on the private network. the iptables_nat module has to be loaded and routing has to be enabled for port forwarding to work.16Harrison_ch14.conf example in this # chapter. #--------------------------------------------------------------echo 1 > /proc/sys/net/ipv4/ip_forward #--------------------------------------------------------------# Allow port forwarding for traffic destined to port 80 of the # firewall’s IP address to be forwarded to port 8080 on server # 192. Here the combination of the firewall’s single IP address. Included only as a reminder. Routing too must be allowed in iptables with the FORWARD chain. Port forwarding is handled by the PREROUTING chain of the nat table. As in masquerading. this includes all NEW inbound connections from the Internet matching the port forwarding port plus all future packets related to the ESTABLISHED connection in both directions.Interface eth0 is the internet interface . #--------------------------------------------------------------# Load the NAT module # # Note: It is best to use the /etc/rc.200 # # . # The external IP address could just be hard coded (“typed in # normally”) #--------------------------------------------------------------external_int=”eth0” external_ip=”`ifconfig $external_int | grep ‘inet addr’ | \ awk ‘{print $2}’ | sed -e ‘s/. This value will not be retained in the # /etc/sysconfig/iptables file.

Interface eth1 is the private network interface #--------------------------------------------------------------iptables -t nat -A PREROUTING -p tcp -i eth0 -d $external_ip \ --dport 80 --sport 1024:65535 -j DNAT --to 192.1.1. Also notice how you have to specify a separate -m multiport option whenever you need to match multiple nonsequential ports for both source and destination. Because the firewall has more than one IP address. In this example.200 \ --dport 8080 --sport 1024:65535 -m state --state NEW -j ACCEPT iptables -A FORWARD -t filter -o eth0 -m state \ --state NEW. not just to a particular port.253.168.ESTABLISHED.26.168. the packets are routed via the filter table’s # FORWARD chain. Instead. only connections on ports 80.158. I can’t recommend MASQUERADE.1. is translated to a single server on the protected subnet.1.100 to 102).29. . and 22 are allowed through by the FORWARD chain.qxd 2/25/05 10:06 AM Page 232 232 Linux Firewalls Using iptables Chapter 14 # . use SNAT to specify the alias IP address to be used for connections initiated by all other servers in the protected network.168.1.158.RELATED -j ACCEPT iptables -A FORWARD -t filter -i eth0 -m state \ --state ESTABLISHED. #--------------------------------------------------------------iptables -A FORWARD -p tcp -i eth0 -o eth1 -d 192.253. # Connections on port 80 to the target machine on the private # network must be allowed. it will force masquerading as the IP address of the primary interface and not as any of the alias IP addresses the firewall may have. This is different from masquerading.168. all traffic to a particular public IP address.0 home network in which all the servers appear on the Internet as IP address 97.16Harrison_ch14.168.RELATED -j ACCEPT Static NAT In this example. ☞ Creates a many to one NAT for the 192. Note Although the nat table translates all traffic to the target servers (192. the firewall: ☞ Uses one to one NAT to make the server 192. 443.200:8080 #--------------------------------------------------------------# After DNAT.100 on your home network appear on the Internet as IP addresses 97.

158. #--------------------------------------------------------------# Load the NAT module # # Note: It is best to use the /etc/rc.qxd 2/25/05 10:06 AM Page 233 Sample iptables Scripts 233 You will have to create alias IP addresses for each of these Internet IPs for one to one NAT to work.158.1.100 # POSTROUTING statements for 1:1 NAT .Inbound) # Anywhere 192.100 97. Included only as a reminder. #--------------------------------------------------------------echo 1 > /proc/sys/net/ipv4/ip_forward #--------------------------------------------------------------# NAT ALL traffic: ########### # REMEMBER to create aliases for all the internet IP addresses below ########### # # TO: FROM: MAP TO SERVER: # 97.16Harrison_ch14.local example in this # chapter.100 (1:1 NAT .168.253.26 (1:1 NAT .conf example in this # chapter.158.26 -i eth0 \ -j DNAT --to-destination 192. Included only as a reminder. Frequently used to NAT # connections from the Internet to your home network # # .253.168.168. This value will not be retained in the # /etc/sysconfig/iptables file.29 (FW IP) # # SNAT is used to NAT all other outbound connections initiated # from the protected network to appear to come from # IP address 97.253.0/24 97.253.29 # # POSTROUTING: # NATs source IP addresses.1. #--------------------------------------------------------------modprobe iptable_nat #--------------------------------------------------------------# Enable routing by modifying the ip_forward /proc filesystem file # # Note: It is best to use the /etc/sysctl.26 Anywhere 192.168.Outbound) # Anywhere 192.Interface eth0 is the internet interface # .158.1. This value will not be retained in the # /etc/sysconfig/iptables file.1.158. Frequently used to NAT connections from # your home network to the Internet # # PREROUTING: # NATs destination IP addresses.253.Interface eth1 is the private network interface #--------------------------------------------------------------# PREROUTING statements for 1:1 NAT # (Connections originating from the Internet) iptables -t nat -A PREROUTING -d 97.

16Harrison_ch14. ☞ Automatically writes an entry to the /var/log/messages file and then executes the next rule.443.1.1.1. therefore.168.RELATED -j ACCEPT TROUBLESHOOTING IPTABLES A number of tools are at your disposal for troubleshooting iptables firewall scripts.253.158.100 \ -m multiport --dport 80. If you .26 # POSTROUTING statements for Many:1 NAT # (Connections originating from the entire home network) iptables -t nat -A POSTROUTING -s 192.168.ESTABLISHED.RELATED -j ACCEPT # Allow forwarding for all 1:1 NAT connections originating on # the Internet that have already passed through the NEW forwarding # statements above iptables -A FORWARD -t filter -i eth0 -m state \ --state ESTABLISHED.22 \ -m state --state NEW -j ACCEPT # Allow forwarding for all New and Established SNAT connections # originating on the home network AND already established # DNAT connections iptables -A FORWARD -t filter -o eth0 -m state \ --state NEW.qxd 2/25/05 10:06 AM Page 234 234 Linux Firewalls Using iptables Chapter 14 # (Connections originating from the home network servers) iptables -t nat -A POSTROUTING -s 192.158. Notice how you # use the real IP addresses here) iptables -A FORWARD -p tcp -i eth0 -o eth1 -d 192.168. you have to add a matching rule with a DROP target immediately after the LOG rule.0/24 \ -j SNAT -o eth1 --to-source 97.29 # Allow forwarding to each of the servers configured for 1:1 NAT # (For connections originating from the Internet. You should be aware that the LOG target: ☞ Logs all traffic that matches the iptables rule in which it is located.100 -o eth0 \ -j SNAT --to-source 97. If you want to log only unwanted traffic. Checking the Firewall Logs You track packets passing through the iptables list of rules using the LOG target. One of the best methods is to log all dropped packets to the /var/log/messages file.253.

because by default iptables doesn’t state why the packet was logged in its log message. “Simple Network Troubleshooting.200. and NAT related statements.168.113 LEN=76 TOS=0x10 PREC=0x00 TTL=64 ID=0 DF PROTO=UDP SPT=123 DPT=123 LEN=56 The traffic in all these examples isn’t destined for the firewall.168.30 DST=192. .16Harrison_ch14.93. This example logs a summary of failed packets to the file /var/log/messages.42. OUTPUT. #--------------------------------------------------------------# Log and drop all other packets to file /var/log/messages # Without this we could be crawling around in the dark #--------------------------------------------------------------iptables -A OUTPUT -j LOG iptables -A INPUT -j LOG iptables -A FORWARD -j LOG iptables -A OUTPUT -j DROP iptables -A INPUT -j DROP iptables -A FORWARD -j DROP Here are some examples of the output of this file: ☞ Firewall denies replies to DNS queries (UDP port 53) destined to server 192. you should check your INPUT.” to determine whether the data is reaching your firewall at all and.1.1. then follow the steps in Chapter 4.102 DST=207.81.168. therefore.102 on the home network Feb 23 20:33:50 bigboy kernel: IN=wlan0 OUT= MAC=00:06:25:09:69:80:00:a0:c5:e1:3e:88:08:00 SRC=192. If the firewall’s IP address is involved.255 LEN=241 TOS=0x00 PREC=0x00 TTL=64 ID=0 DF PROTO=UDP SPT=138 DPT=138 LEN=221 ☞ Firewall denies Network Time Protocol (NTP UDP port 123) Feb 23 20:58:48 bigboy kernel: IN= OUT=wlan0 SRC=192. then you should focus on the INPUT and OUTPUT statements If nothing shows up in the logs.100 DST=192.1.168.102 LEN=220 TOS=0x00 PREC=0x00 TTL=54 ID=30485 PROTO=UDP SPT=53 DPT=32820 LEN=200 ☞ Firewall denies Windows NetBIOS traffic (UDP port 138) Feb 23 20:43:08 bigboy kernel: IN=wlan0 OUT= MAC=ff:ff:ff:ff:ff:ff:00:06:25:09:6a:b5:08:00 SRC=192. if it is not.qxd 2/25/05 10:06 AM Page 235 Troubleshooting iptables 235 don’t.1. the location on your network that could be causing the problem. you’ll find yourself logging both desired and unwanted traffic with no way of discerning between the two.1.168. FORWARD. You can use the contents of this file to determine which TCP/UDP ports you need to open to provide access to specific traffic that is currently stopped.

qxd 2/25/05 10:06 AM Page 236 236 Linux Firewalls Using iptables Chapter 14 As a general rule. but a firewall is never sufficient. . restricted and monitored physical access to your servers. secured cabling. Unfortunately. probably most importantly. running the service iptables save command before restarting won’t help either. reliable power and cooling. you won’t be able to access the public NAT IP addresses from servers on your home network. You can then use the logging output in /var/log/messages to make sure that the translations are occurring correctly and iptables isn’t dropping the packets after translation occurs. then you will face this problem. and it is well worth the money to invest in and learn from a book that specializes in the topic.d/iptables script running without the typical [OK] or [FAILED] messages. well trained and motivated employees. and. Basic NAT testing requires you to ask a friend to try to connect to your home network from the Internet. regular application patching. If you have just installed iptables and have never applied a policy. then symptoms include the firewall status always being stopped and the /etc/init. redundant hardware. Web site security involves not just protection from corrupted packets or maliciously overwhelming volumes of traffic. enforced password policies. You have to create this file: [root@bigboy tmp]# service iptables start [root@bigboy tmp]# [root@bigboy tmp]# touch /etc/sysconfig/iptables [root@bigboy tmp]# chmod 600 /etc/sysconfig/iptables [root@bigboy tmp]# service iptables start Applying iptables firewall rules: [ OK ] [root@bigboy tmp]# CONCLUSION A firewall is a critical part of any establishment that connects to an unprotected network such as the Internet. If none exists. Security should be viewed as anything that contributes to the desired risk-free functioning of your site.16Harrison_ch14. but also involves daily data backups to help recovery from device failures. iptables Won’t Start The iptables startup script expects to find the /etc/sysconfig/iptables before it starts.

and. FTP OVERVIEW FTP relies on a pair of TCP ports to get the job done. Most Web-based download sites use the built-in FTP capabilities of Web browsers. It operates using two connection channels: 237 .conf File FTP Security Issues Troubleshooting FTP Tutorial Conclusion The File Transfer Protocol (FTP) is one of the most common means of copying files between servers over the Internet.17Harrison_ch15. Linux is no exception.qxd 2/25/05 10:06 AM Page 237 C H A P T E R 15 Linux FTP Server Setup IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ FTP Overview Problems with FTP and Firewalls How to Download and Install VSFTPD How to Get VSFTPD Started Testing the Status of VSFTPD The vsftpd. most serveroriented operating systems usually include an FTP server application as part of the software suite. This chapter will show you how to convert your Linux box into an FTP server using the default Very Secure FTP Daemon (VSFTPD) package included in Fedora. therefore.

the connection is initiated from the FTP client. Active FTP FTP Control Connection to Port 21 from High Port High Port Client Computer FTP Data Connection Initiation from Port 20 on Server to High Port on Client Server Port 20 Passive FTP FTP Control Connection to Port 21 from High Port High Port Client Computer Server High Port FTP Data Connection Initiation from High Port on Client to High Port on the Server Figure 15. as well as the FTP server’s responses to those commands.17Harrison_ch15. there are two additional types of FTP: regular FTP. in which general access is provided to the FTP server using a well known universal login method.1 Active and passive FTP. From a user management perspective. in which files are transferred using the username and password of a regular user FTP server. . For passive FTP.1. and anonymous FTP. the two main types of FTP are active and passive. there are several varieties of FTP. but any data sent back (such as ls directory lists or actual file data in either direction) will go over the data connection. the FTP server initiates a data transfer connection back to the client. go over the control connection.qxd 2/25/05 2:40 PM Page 238 238 Linux FTP Server Setup Chapter 15 ☞ FTP control channel. Types of FTP From a networking perspective. These are illustrated in Figure 15. TCP Port 21: All commands you send. ☞ FTP data channel. In addition to these channels. Take a closer look at each type. TCP Port 20: This port is used for all subsequent data transfers between the client and server. In active FTP.

Your client connects to the FTP server by establishing an FTP control connection to port 21 of the server. Passive FTP Passive FTP works differently: 1. Thus the ls listing that you asked for comes back over the port 20 to high port connection. Whenever the client requests data over the control connection. passive FTP works better for clients protected by a firewall. Active FTP may fail in cases where the client is protected from the Internet via many to one NAT (masquerading). because the firewall will not know which of the many servers behind it should receive the return connection. Whenever the client requests data over the control connection. Because the client always initiates the required connections. Passive FTP should be viewed as the server never making an active attempt to connect to the client for FTP data transfers. not the port 21 control connection. The source port of these data transfer connections is always a high port on the client with a destination port of a high port on the server. you’ll probably have to accommodate both forms when deciding upon a security policy for your FTP server. transfers data in a counter intuitive way to the TCP standard.17Harrison_ch15. 2. 2. Your client connects to the FTP server by establishing an FTP control connection to port 21 of the server. the server initiates data transfer connections back to the client. the client initiates the data transfer connections to the server. FTP active mode. . Regular FTP By default. As Windows defaults to active FTP and Linux defaults to passive.qxd 2/25/05 2:40 PM Page 239 FTP Overview 239 Active FTP The sequence of events for active FTP is: 1. Your commands such as ls and get are sent over that connection. and the destination port is a high port (greater than 1024) on the client. therefore. The source port of these data transfer connections is always port 20 on the server. 3. the VSFTPD package allows regular Linux users to copy files to and from their home directories with an FTP client using their Linux usernames and passwords as their login credentials. Your commands such as ls and get are sent over this connection. as it selects port 20 as its source port (not a random high port that’s greater than 1024) and connects back to the client on a random high port that has been pre-negotiated on the port 21 control connection.

The reason is that the firewall is blocking the return connection from the server to the client (from port . As seen in Chapter 6.” using anonymous FTP as a remote user is fairly straightforward. Client Protected by a Firewall Problem Typically firewalls don’t allow any incoming connections at all. Scripts. Common uses include downloading software updates and MP3s and uploading diagnostic information for a technical support engineers’ attention. “Installing RPM Software. You have a choice of methods to overcome this. Unlike regular FTP where you login with a preconfigured Linux username and password. PROBLEMS WITH FTP AND FIREWALLS FTP frequently fails when the data has to pass through a firewall. With this type of FTP failure. anonymous FTP requires only a username of anonymous and your e-mail address for the password. The connection then appears to hang. Once logged into a VSFTPD server.” contains examples of how to configure the VSFTPD Linux firewall to function with both active and passive FTP. the active FTP connection appears to work when the client initiates an outbound connection to the server on port 21. or get commands. however. Anonymous FTP Anonymous FTP is the choice of Web sites that need to exchange files with numerous unknown remote users. Anonymous FTP allows you to avoid this difficulty. enabling you to restrict the addition of new files to your system to authorized personnel. VSFTPD can be configured to support user-based and/or anonymous FTP in its configuration file. and Configurations. “Codes. dir. as you’ll see later. because firewalls are designed to limit data flows to predictable TCP ports and FTP uses a wide range of unpredictable TCP ports.17Harrison_ch15. The disadvantage of regular FTP is that it isn’t suitable for general download distribution of software as everyone either has to get a unique Linux user account or has to use a shared username and password. Note Appendix II.qxd 2/25/05 10:06 AM Page 240 240 Linux FTP Server Setup Chapter 15 VSFTPD also has the option of allowing this type of access to only a group of Linux users. as soon as you use the ls. you automatically have access to only the default anonymous FTP directory (/var/ftp in the case of VSFTPD) and all its subdirectories. which frequently blocks active FTP from functioning.

This rule may not be necessary in all cases.2 outlines the general rules needed to allow FTP servers through a firewall. If a firewall allows all outbound connections to the Internet. then passive FTP clients behind a firewall will usually work correctly as the clients initiate all the FTP connections.17Harrison_ch15. server. When an incorrectly configured firewall protects an FTP server. firewalls don’t let any connections come in at all. you may want to allow all Internet users to have access. Server Protected by a Firewall Problem Typically. Solution Table 15.qxd 2/25/05 10:06 AM Page 241 Problems with FTP and Firewalls 241 20 on the server to a high port on the client). 3 Many home-based firewall routers automatically allow traffic for already established connections. Table 15. . High 20 High High New Established3 New Established3 2 In some cases. Solution Table 15. the FTP connection from the client doesn’t appear to work at all for both active and passive FTP.1 shows the general rules you’ll need to allow FTP clients through a firewall. or network. not just a specific client.1 Method Client Protected by Firewall: Required Rules for FTP Destination Port 21 High Connection Type New Established3 Source Source Destination Address Port Address Allow outgoing control connections to server Control FTP client/ High1 FTP server2 channel network FTP server2 21 FTP client/ network Allow the client to establish data channels to remote server Active FTP server2 20 FTP client/ FTP network FTP client/ High FTP server2 network Passive FTP client/ High FTP server2 FTP network FTP server2 High FTP client/ network 1 Greater than 1024.

qxd 2/25/05 10:06 AM Page 242 242 Linux FTP Server Setup Chapter 15 Table 15. HOW TO DOWNLOAD AND INSTALL VSFTPD Most Red Hat and Fedora Linux software products are available in the RPM format. not just a specific client. as in vsftpd-1. or network. 3 Many home-based firewall routers automatically allow traffic for already established connections. 2 Greater than 1024.17Harrison_ch15. If you need a refresher. Chapter 6 covers how to do this in detail.i386. or restart VSFTPD after booting using these commands: [root@bigboy tmp]# service vsftpd start [root@bigboy tmp]# service vsftpd stop [root@bigboy tmp]# service vsftpd restart . remember that the VSFTPD RPM’s filename usually starts with the word “vsftpd” followed by a version number. HOW TO GET VSFTPD STARTED You can start.2. This rule may not be necessary in all cases.rpm. Downloading and installing RPMs isn’t hard. server.1-5. It is best to use the latest version of VSFTPD. When searching for the file.2 Method Server Protected by Firewall: Required Rules for FTP Destination Port 21 High Connection Type New Established3 Source Source Destination Address Port Address Allow incoming control connections to server FTP server Control FTP client/ High2 1 channel network FTP server 21 FTP client/ network1 Allow server to establish data channel to remote client Active FTP server 20 FTP client/ FTP network1 FTP client/ High FTP server network1 Passive FTP client/ High FTP server FTP network1 FTP server High FTP client/ network1 High 20 High High New Established3 New Established3 1 In some cases. you may want to allow all Internet users to have access. stop.

TFTP. This file uses a number of default settings you need to know about: ☞ VSFTPD runs as an anonymous FTP server: Unless you want any remote user to log into to your default FTP directory using a username of anonymous and a password that’s the same as their e-mail address. and XINETD.” You can find a full description of how to configure these versions of Linux for VSFTPD in Appendix III. THE VSFTPD. use the chkconfig command: [root@bigboy tmp]# chkconfig vsftpd on Note In Red Hat Linux version 8. You can set the configuration file’s anonymous_enable directive to no to disable this feature. not uploads from them: You can change this by modifying the anon_upload_enable directive shown later. which lists all the TCP and UDP ports on which the server is listening for traffic. “Fedora Version Differences. I suggest turning this off. there would be no output at all.conf configuration file only when it starts. You’ll also need to simultaneously enable local users to be able to log in by removing the comment symbol (#) before the local_enable instruction. .conf File 243 To configure VSFTPD to start at boot. so you’ll have to restart VSFTPD each time you edit the file in order for the changes to take effect. which is covered in Chapter 16. ☞ VSFTPD allows only anonymous FTP downloads to remote users.0 and earlier. VSFTPD operation is controlled by the xinetd process. “TELNET.” TESTING THE STATUS OF VSFTPD You can always test whether the VSFTPD process is running by using the netstat -a command.qxd 2/25/05 10:06 AM Page 243 The vsftpd.CONF FILE VSFTPD reads the contents of its vsftpd.17Harrison_ch15. This example shows the expected output: [root@bigboy root]# netstat -a | grep ftp tcp 0 0 *:ftp [root@bigboy root]# *:* LISTEN If VSFTPD wasn’t running.

you will # obviously need to create a directory writable by the FTP user. # (Needed even if you want local users to be able to upload files) write_enable=YES # Uncomment to allow the anonymous FTP user to upload files. . ☞ VSFTPD logs FTP access to the /var/log/vsftpd. The configuration file is fairly straightforward as you can see in the snippet: # Allow anonymous FTP? anonymous_enable=YES # Uncomment this to allow local users to log in. You run the risk of filling up your /var partition if you use the default setting. remove or add the # at the beginning of the appropriate line. local_enable=YES # Uncomment this to enable any form of FTP write command. ☞ VSFTPD expects files for anonymous FTP to be placed in the /var/ftp directory: You can change this by modifying the anon_root directive. This only # has an effect if global write enable is activated. (Default = /var/ftp) #anon_root=/data/directory To activate or deactivate a feature. xferlog_enable=YES # You may override where the log file goes if you like. It is best to make the anonymous FTP directory reside in its own dedicated partition. #anon_mkdir_write_enable=YES # Activate logging of uploads/downloads.qxd 2/25/05 10:06 AM Page 244 244 Linux FTP Server Setup Chapter 15 ☞ VSFTPD doesn’t allow anonymous users to create directories on your FTP server: You can change this by modifying the anon_mkdir_write_enable directive. # The default is shown# below. Also. #anon_upload_enable=YES # Uncomment this if you want the anonymous FTP user to be able to create # new directories.17Harrison_ch15.log log file: You can change this by modifying the xferlog_file directive. There is always the risk with anonymous FTP that users will discover a way to write files to your anonymous FTP directory.log # The directory which vsftpd will try to change # into after an anonymous login. #xferlog_file=/var/log/vsftpd.

ftpusers file. FTP logins and data transfers are not encrypted. though very convenient. The VSFTPD package creates this file with a number of entries for privileged users that normally shouldn’t have FTP access.conf Options There are many other options you can add to this file: ☞ Limiting the maximum number of client connections (max_clients) ☞ Limiting the number of connections by source IP address (max_per_ip) ☞ Setting the maximum rate of data transfer per anonymous login (anon_max_rate) ☞ Setting the maximum rate of data transfer per non-anonymous login (local_max_rate) Descriptions on this and more can be found in the pages.conf man FTP SECURITY ISSUES FTP has a number of security drawbacks. vsftpd.17Harrison_ch15. thereby increasing the risk of data or passwords being compromised. and you can change the configuration to not display the FTP server’s software version information. then you should create a write-only directory within /var/ftp/pub. it is a good idea to let these entries remain and add new entries for additional security. You can restrict an individual Linux user’s access to non-anonymous FTP. but you can overcome them in some cases. This will allow your users to upload but not access other files uploaded by other users. you may restrict FTP access to certain users by adding them to the list of users in the /etc/vsftpd. Anonymous Upload If you want remote users to write data to your FTP server.ftpusers File For added security. As FTP doesn’t encrypt passwords. The commands you need are: [root@bigboy tmp]# mkdir /var/ftp/pub/upload [root@bigboy tmp]# chmod 722 /var/ftp/pub/upload . but unfortunately. The /etc/vsftpd.qxd 2/25/05 10:06 AM Page 245 FTP Security Issues 245 Other vsftpd.

because you run the risk of accidentally allowing unknown persons to upload files to your server. which in turn can clog your server’s Internet access and drive up your bandwidth charges. Secure Copy (SCP) and Secure FTP (SFTP) provide encryption and could be considered as an alternative to FTP for trusted users. Typical symptoms of this are either connection timeouts or the ability to use the ls command to view the contents of a directory without the ability to either upload or download files. This sort of unintended activity can quickly fill up your hard drive with illegal software. The directive in this file is: ftpd_banner= New Banner Here Using SCP as Secure Alternative to FTP One of the disadvantages of FTP is that it does not encrypt your username and password.” TUTORIAL FTP has many uses. one of which is allowing numerous unknown users to download files.conf file to make it harder for malicious users to determine the type of system you have. images. “Simple Network Troubleshooting.17Harrison_ch15.qxd 2/25/05 10:06 AM Page 246 246 Linux FTP Server Setup Chapter 15 FTP Greeting Banner Change the default greeting banner in the vsftpd. SCP does not support anonymous services. The most common sources of day-to-day failures are incorrect usernames and passwords. Initial setup failures could be caused by firewalls along the path between the client and server blocking some or all types of FTP traffic. and music for the world to download. Follow the firewall rule guidelines to help overcome this problem. Connection problems could also be the result of typical network issues outlined in Chapter 4. however. You have to be careful. . a feature that FTP does support. This could make your user account vulnerable to an unauthorized attack from a person eavesdropping on the network connection. TROUBLESHOOTING FTP You should always test your FTP installation by attempting to use an FTP client to log into your FTP server to transfer sample files.

use /home/ftpusers and a user group name of ftp-users for the remote users: [root@bigboy tmp]# groupadd ftp-users [root@bigboy tmp]# mkdir /home/ftp-docs 5. Start VSFTP. and make their default directory /home/ftp-docs: [root@bigboy tmp]# useradd -g ftp-users -d /home/ftp-docs user1 [root@bigboy tmp]# useradd -g ftp-users -d /home/ftp-docs user2 [root@bigboy tmp]# useradd -g ftp-users -d /home/ftp-docs user3 [root@bigboy tmp]# useradd -g ftp-users -d /home/ftp-docs user4 [root@bigboy tmp]# passwd user1 [root@bigboy tmp]# passwd user2 [root@bigboy tmp]# passwd user3 [root@bigboy tmp]# passwd user4 7. Here are the steps: 1. Change the permissions of the files in the /home/ftp-docs directory to read-only access by the group: [root@bigboy tmp]# chown root:ftp-users /home/ftp-docs/* [root@bigboy tmp]# chmod 740 /home/ftp-docs/* .conf file: # Allow anonymous FTP? # anonymous_enable=YES anonymous_enable line in the 2. but a group of trusted users need to have read-only access to a directory for downloading files.17Harrison_ch15. anonymous FTP is not desired.qxd 2/25/05 10:06 AM Page 247 Tutorial 247 FTP Users with Read-Only Access to a Shared Directory In this example. Enable individual logins by making sure you have the uncommented in the vsftpd. Make the directory accessible to the ftp-users group: [root@bigboy tmp]# chmod 750 /home/ftp-docs [root@bigboy tmp]# chown root:ftp-users /home/ftp-docs 6. Disable anonymous FTP. 8. [root@bigboy tmp]# service vsftpd start 4. Copy files to be downloaded by your users into the /home/ftp-docs directory. In this case. Comment out the vsftpd. local_enable=YES local_enable line 3.conf file: # Uncomment this to allow local users to log in. Create a user group and shared directory. Add users.

dude (vsFTPd 1. Sample Login Session to Test Functionality Here is a simple test procedure you can use to make sure everything is working correctly: 1.168. ftp> But we can view and download a copy of the VSFTPD RPM on the FTP server bigboy: ftp> ls 227 Entering Passive Mode (192.100 (192.1.1.100 Connected to 192.100:root): user1 331 Please specify the password.1. [root@smallfry tmp]# ll total 1 -rw-r--r-. Connect to Bigboy via FTP: [root@smallfry tmp]# ftp 192. you must restart VSFTPD for the configuration file changes to take effect. Password: 230 Login successful.168.100) 220 ready.168. Remote system type is UNIX. -rwxr----.1. break me) Name (192.0: beat me. Have fun.1.100.rpm .1.168.168. then you should set the write_enable line in your vsftpd.qxd 2/25/05 10:06 AM Page 248 248 Linux FTP Server Setup Chapter 15 Users should now be able to log in via FTP to the server using their new usernames and passwords.168.17Harrison_ch15.0-1.1. Check for the presence of a test file on the FTP client server.i386.conf file to no: write_enable = NO Remember.100.1 0 502 76288 Jan 04 17:06 vsftpd-1.173) 150 Here comes the directory listing. If you absolutely don’t want any FTP users to be able to write to any directory.181. you can’t do an upload transfer of testfile to bigboy: ftp> put testfile local: testfile remote: testfile 227 Entering Passive Mode (192.35.210) 553 Could not create file. ftp> As expected.1.1 root root 0 Jan 4 09:08 testfile [root@smallfry tmp]# 2. Using binary mode to transfer files.

tmp local: vsftpd-1. CONCLUSION FTP is a very useful software application that can have enormous benefit to a Web site or to collaborative computing in which files need to be shared between business partners. Password: 530 Login incorrect.1. because FTP clients are a part of all operating systems and Web browsers.1.i386. Login failed. ftp> quit 221 Goodbye.168.1.01.1.rpm 227 Entering Passive Mode (192.rpm.100 Connected to 192.1. it is universally accessible.100.i386. “Secure Remote Logins and File Copying.168.168.0-1.156) 150 Opening BINARY mode data connection for vsftpd-1.100 (192.rpm.499 secs (1.1.” .i386. 76288 bytes received in 0.168.i386. 226 File send OK.1.17Harrison_ch15. you can make this a regular part of your FTP server’s operation.rpm vsftpd-1.1.rpm (76288 bytes).5e+02 Kbytes/sec) ftp> exit 221 Goodbye.1.0: beat me.i386.168. [root@smallfry tmp]# Now that testing is complete. Although insecure.0-1. If data encryption security is of great importance to you.01.100) 220 ready.tmp remote: vsftpd-1. break me) Name (192. You can find more information on it in Chapter 17.44. [root@smallfry tmp]# As expected. ftp> get vsftpd-1.1. dude (vsFTPd 1.qxd 2/25/05 10:06 AM Page 249 Conclusion 249 226 Directory send OK. then you should probably consider SCP as a possible alternative.100:root): anonymous 331 Please specify the password.0-1.1. an anonymous FTP fails: [root@smallfry tmp]# ftp 192.

17Harrison_ch15.qxd 2/25/05 10:06 AM Page 250 .

instead they often offload a lot of this work to a program suite made just for this purpose: xinetd.d directory control the starting and stopping of the xinetd daemon.qxd 2/25/05 10:07 AM Page 251 C H A P T E R 16 TELNET. Controlling xinetd The scripts in the /etc/init. TFTP. so day-to-day xinetd operation is frequently limited to only starting and stopping xinetd-managed applications. You can start. and restart xinetd after booting by using: [root@bigboy tmp]# service xinetd start [root@bigboy tmp]# service xinetd stop [root@bigboy tmp]# service xinetd restart xinetd RPM /etc/xinetd. and xinetd IN THIS CHAPTER ☞ Managing xinetd Programs ☞ TELNET ☞ Conclusion Many network-enabled Linux applications don’t rely on themselves to provide restricted access or bind to a particular TCP port.conf as 251 . MANAGING The XINETD PROGRAMS is installed by default in Fedora Linux and uses its main configuration file. Fortunately you usually don’t have to edit this file.18Harrison_ch16. stop. and chkconfig controls its behavior at boot time.

xinetd configured to start at boot. This governs whether xinetd is allowed to start them or not. you can use the chkconfig [root@bigboy tmp]# chkconfig xinetd on Controlling xinetd-Managed Applications Xinetd-managed applications all store their configuration files in the /etc/xinetd. Each configuration file has a disable statement that you can set to yes or no.18Harrison_ch16. One of the disadvantages of Telnet is that the data is sent as clear text. You don’t have to edit these files to activate or deactivate the application. which uses varying degrees of encryption. TFTP. and xinetd Chapter 16 To get command. Many network devices don’t have SSH clients. This means that it is possible for someone to use a network analyzer to peek into your data packets and see your username and password. The TELNET server RPM is installed and disabled by default on Fedora Linux. Using The TELNET Client The command to do remote logins via TELNET from the command line is simple. I’ll show you how to limit your exposure to TELNET’s insecurities later in this chapter.d directory. the older TELNET application remains popular. The chkconfig command does that for you automatically and also stops or starts the application accordingly too! Here is an example of the activation and deactivation of the Samba SWAT Web GUI management application: [root@bigboy tmp]# chkconfig swat on [root@bigboy tmp]# chkconfig swat off TELNET TELNET is a program that allows users to log into your server and get a command prompt just as if they were logged into the VGA console.qxd 2/25/05 10:07 AM Page 252 252 TELNET. In spite of this. You enter the word “telnet” and then the IP address or server name to which you want to connect. making Telnet the only means of accessing other devices and servers from them. A more secure method for remote logins would be via Secure Shell (SSH). .

i386. Most Red Hat and Fedora Linux software products are available in the RPM format.0. The user looks at the routing table and then logs out. Chapter 6.my-web-site.0.0.” covers how to do this in detail.4.0.org) (10:35 on Sunday.0 255. If you need a refresher. [root@bigboy tmp]# telnet 192.105 Trying 192.1 0. as in telnet-server-0.18Harrison_ch16.. Connected to 192.0.0.17-28.168.0 255.255 0.255.255. Linux 2. remember that the TELNET server RPM’s filename usually starts with telnet-server followed by a version number. [root@bigboy tmp]# MSS Window irtt 40 40 40 40 0 0 0 0 0 0 0 0 Installing the TELNET Server Software Older versions of Red Hat had the TELNET server installed by default.0 255.0.0.0 UG wlan0 [peter@smallfry peter]$ exit logout Connection closed by foreign host.0 U lo 0.1.0.255 UH wlan0 192.0.rpm.1.105. .168.0. Escape character is ‘^]’. “Installing RPM Software.0 0.105.1. 05 January 2003) Login: peter Password: Last login: Fri Nov 22 23:29:44 on ttyS0 You have new mail. Downloading and installing RPMs isn’t hard.0 192.0.255.qxd 2/25/05 2:41 PM Page 253 TELNET 253 Here is an example of someone logging into a remote server named smallfry from server Bigboy. and you will have to install it yourself.168.0.1. Fedora Linux doesn’t do this..168.0 U wlan0 127.168.0 0.255.0.18-14 (smallfry. [peter@smallfry peter]$ [peter@smallfry peter]$ netstat –nr Kernel IP routing table Destination Gateway Genmask Flags Iface 255.255. When searching for the file.255.1.

good port scanning programs can detect TELNET and other applications running on alternative ports. Call it # Local services stelnet 7777/tcp telnet # “secure” . and add an entry for a new service.18Harrison_ch16. then there will be no response. and xinetd Chapter 16 Setting Up a TELNET Server To set up a TELNET server. you should also try to ensure that TELNET sessions run over secure internal networks or across VPNs to reduce the risk of exposing sensitive data to unauthorized eyes. even after the next reboot: [root@bigboy tmp]# chkconfig telnet off Basic TELNET Security There are a number of things you can do to improve the security of TELNET. Remember that this isn’t a foolproof strategy. TFTP.qxd 2/25/05 10:07 AM Page 254 254 TELNET. but it makes it less likely to be detected as TELNET traffic. [root@bigboy tmp]# netstat -a | grep telnet tcp 0 0 *:telnet LISTEN [root@bigboy tmp]# *:* You can also use the chkconfig will be started on the next reboot: --list command to verify that TELNET [root@bigboy tmp]# chkconfig --list | grep telnet telnet: on [root@bigboy tmp]# Stopping a TELNET Server Use the chkconfig command to deactivate TELNET. /etc/services file. For example. which is used to check the TCP/UDP ports on which your server is listening. Check out some other options. If it isn’t running. Edit your stelnet. use the chkconfig command to activate TELNET: [root@bigboy tmp]# chkconfig telnet on You can test whether the TELNET process is running with the netstat command. Let Telnet Listen on Another TCP Port Letting TELNET run on an alternate TCP port doesn’t encrypt the traffic. The steps are: 1.

Make the new service stelnet.6. service stelnet { flags = REUSE socket_type = stream wait = no user = root server = /usr/sbin/in.521 on an i686 login: Let TELNET Allow Connections from Trusted Addresses You can restrict TELNET logins’ access to individual remote servers by using the only_from keyword in the TELNET configuration file.168. This is done using the telnet command with the TCP port as the second argument.d/stelnet file.0:* [root@bigboy tmp]# LISTEN You should now be able to log into the new stelnet server on port 7777. Edit the new /etc/xinetd.d/stelnet 3. Fedora Core release 2 (Tettnang) Kernel 2.d/telnet and [root@bigboy tmp]# cp /etc/xinetd. Copy the TELNET configuration file called call it /etc/xinetd.0. Escape character is ‘^]’.0:7777 0.100.1. Add a list of trusted servers to the /etc/xinetd.100 7777 Trying 192.qxd 2/25/05 10:07 AM Page 255 TELNET 255 2.168. and add a port statement for TCP port 7777: # default: on # description: The telnet server serves telnet sessions # unencrypted username/password pairs for authentication.d/telnet /etc/xinetd.1. Check to make sure your server is now listening on port 7777 with the netstat command: [root@bigboy tmp]# netstat -an | grep 777 tcp 0 0 0.8-1.d/telnet file separated by spaces: # default: on # description: The telnet server serves telnet sessions .telnetd log_on_failure += USERID disable = no port = 7777 } 4..18Harrison_ch16.1..d/stelnet: /etc/xinetd. [root@smallfry tmp]# telnet 192.0.0.100. Connected to 192.168.0. Here’s how: 1. Use chkconfig to activate stelnet: [root@bigboy tmp]# chkconfig stelnet on 5.

1.1. The provided TFTP examples use equipment from Cisco Systems.168.telnetd log_on_failure += USERID disable = no only_from = 192. they can be used to upload even partial configurations.1 192.200 } 2.168. a leading networking hardware manufacturer. TFTP servers can be used to upload new configurations to replacement devices after serious hardware failures. Connected to 192. Finally.100 127.100 Trying 192. such as files containing updated access control lists (ACLs) that restrict access to networks and even the regular application of new passwords. . but it will become increasingly important in an expanding small office/home office (SOHO) environment.qxd 2/25/05 10:07 AM Page 256 256 TELNET..18Harrison_ch16. Escape character is ‘^]’. TFTP can be used with great versatility as a network management tool and not just for saving files. TFTP may not be an application used regularly in a home. service telnet { flags = REUSE socket_type = stream wait = no user = root server = /usr/sbin/in. Servers that are not on the trusted list get the message “Connection closed by foreign host.100.0.100.1.168.1. Restart TELNET: [root@bigboy tmp]# chkconfig telnet off [root@bigboy tmp]# chkconfig telnet on 3.168.1.. Connection closed by foreign host. Test the TELNET session.” [root@smallfry tmp]# telnet 192.0. They also can be used for uploading new versions of software to be run as network devices. TFTP.168. which is why the topic is covered here. and xinetd Chapter 16 # unencrypted username/password pairs for authentication. [root@smallfry tmp]# TFTP Many networking equipment manufacturers allow you to back up live configurations of their devices to centralized servers via the TFTP protocol.

This example creates a new tftpboot directory in the /var partition.33-3. TFTP files of increasing size could eventually fill the partition. as in: tftp-server-0. affecting your ability to install new software or even the overall performance of your system. remember that the TFTP server RPM’s filename usually starts with word tftp-server and is followed by a version number. Here’s an example of what to do for a SOHO firewall named pixfw and a configuration filename that matches Cisco’s standard naming scheme of devicename-config: [root@bigboy [root@bigboy [root@bigboy total 1631 -rw-rw-rw. see Chapter 6. If you need a refresher.rpm. or create your own directory just for this purpose and create a /tftpboot symbolic link to it. and then creates a symbolic link that makes this directory appear to also be the /tftpboot directory: [root@bigboy tmp]# mv /tftpboot /var [root@bigboy tmp]# ln -s /var/tftpboot /tftpboot You must restart xinetd for the new configuration to take effect: [root@bigboy tmp]# chkconfig tftp on Each device must have a configuration file in the /tftpboot directory.1 [root@bigboy tmp]# touch /tftpboot/pixfw-config tmp]# chmod 666 /tftpboot/pixfw-config tmp]# ll /tftpboot/ root root 3011 Oct 29 14:09 pixfw-config tmp]# . By default.d/tftp file via the server_args option. Change this in the /etc/xinetd.qxd 2/25/05 2:42 PM Page 257 TFTP 257 Installing the TFTP Server Software Most Red Hat and Fedora Linux software products are available in the RPM format.18Harrison_ch16.i386. It is usually best to place the TFTP files in a partition other than the root partition. Configuring the TFTP Server The procedure to set up a TFTP Server is straightforward. When searching for the file. Downloading and installing RPMs isn’t hard. the TFTP application expects files to be located in the /tftpboot directory.

18Harrison_ch16. Log onto the device...1. TFTP write ‘/pixfw-config’ at 192. The examples assume that the TFTP server’s IP address is 192.100 /pixfw-config pixfw(config)# exit 2. get into Enable mode. and xinetd Chapter 16 You can test whether the TFTP process is running with the netstat command. [root@bigboy tmp]# netstat -a | grep tftp udp 0 0 *:tftp [root@bigboy tmp]# *:* Saving Cisco Configurations to the TFTP Server You’ll now have to configure your Cisco router/firewall to use the TFTP server. and then enter the TFTP commands to initially configure TFTP: pixfw> enable Password: ******** pixfw# configure terminal pixfw(config)# tftp-server inside 192. Cisco PIX Firewall Follow these steps on a PIX firewall: 1.168. If it isn’t running then there will be no response. you need to log onto the device.168.1. Save the configuration to non-volatile memory: pixfw# write memory Building configuration.100 on interface 1 [OK] pixfw# Your firewall configuration has now been successfully saved for later use in the event of unrecoverable human error or hardware failure.qxd 2/25/05 10:07 AM Page 258 258 TELNET. Save the configuration to the TFTP server: pixfw# write network Building configuration. Cryptochecksum: 3af43873 d35d6f06 51f8c999 180c2342 [OK] pixfw# 3.168. which is used to check the TCP/UDP ports on which your server is listening. Cisco Switch Running CATOS To save the configuration of a Catalyst-series switch running CATOS.. and then enter the write net TFTP command: .. get into Enable mode.100. TFTP.1.

.18Harrison_ch16.168.... get into Enable mode.100:69 . log onto the device and enter the TFTP commands as seen below: ciscocss# copy running-config tftp 192.1. [OK] ciscold# exit . Finished network upload.168. and then enter the TFTP commands: ciscold> ena Password: ciscold# write net 192. and then enter the TFTP commands: ciscorouter> enable ciscorouter# write net Remote host [192.100]? 192..100 (y/n) [n]? y ..168. get into Enable mode. ciscocss# exit Cisco Local Director To save the configuration of a Cisco Local Director load balancer.1.168. log onto the device. writing configuration to //ciscold-config on 192.100 ciscocss-config Working. switch to Configure mode.1.1.168.168..100 Name of configuration file to write [ciscorouter-config]? ciscorouterconfig Write file ciscorouter-config on host 192.168..100] Name of configuration file?[ciscoswitch-config] Upload configuration to ciscoswitch-config on 192...1. IP address or name of remote host? [192.168.1.(\) 100% Connecting (/) Completed successfully.100 ciscold-config Building configuration.qxd 2/25/05 10:07 AM Page 259 TFTP 259 ciscoswitch> (enable) wr net This command shows non-default configurations only.100? [confirm] y ciscorouter# exit Cisco CSS 11000 Arrowpoints To save the configuration of a Cisco CSS-series load balancer.1. (30907 bytes) ciscoswitch> (enable) Cisco Router To save the configuration of a router.1. Use ‘write network all’ to show both default and non-default configurations... switch to Configure mode. log onto the device..

qxd 2/25/05 10:07 AM Page 260 260 TELNET.18Harrison_ch16.100/config.100 (via FastEthernet0/0): !!!!!! [OK . it would be referred to as config.0. it would be referred to as /tftboot/configs/config.0 0.1.168.1.file Configure using tftp://192. a small file containing a new encrypted password and access control list is uploaded from the TFTP server and inserted into a router configuration. Consider this sample configure network command: ciscorouter>ena Password: ciscorouter#configure network Host or network configuration file [host]? This command has been replaced by the command: ‘copy <url> system:/running-config’ Address or name of remote host []? 192.10. and xinetd Chapter 16 Uploading Cisco Configurations from the TFTP Server From time to time you may have to upload configurations from your TFTP server to your network equipment.file.168. and it looks like this: ! ! Set the console password ! line con 0 password 7 $1$qDwqJEjunK$tuff0HE/g31/b7G/IZ ! ! Delete and recreate access list #10 ! no access-list 10 access-list 10 permit 192.26521/52224 bytes] ciscorouter# Here’s a sample copy tftp: running-config command: . So if the file was located in the /tftpboot directory. In both cases.0 0.1. Sample Upload Configuration File The configuration file is named config.168.file from 192. In this example.file.168. you are prompted for the IP address of the TFTP server and the name of the file with the configuration commands.file.0.168.255 access-list 10 permit 192.0.file? [confirm] Loading config. but if it were in the /tftboot/configs directory. TFTP. The filename provided is always relative to the /tftpboot directory.255 Procedure to Upload Configuration File Uploading the file can be done using either the copy tftp: running-config or the older configure network commands.100 Source filename []? config.1.0.

but expect to edit it if the interface names and software versions are different. . 2..168. Using TFTP to Restore Your Router Configuration In disastrous cases. Disconnect the router from the network... Log into the router via the console and verify the configuration is OK.. 5. Use the copy command to copy the backup configuration from the TFTP server to your startup configuration in NVRAM. Reload the router without saving the live running configuration to overwrite the startup configuration. If the replacement unit is identical.file Destination filename [running-config]? Accessing tftp://192. a default route ! and the TFTP setup commands.912 secs (26521 bytes/sec) ciscorouter# Always remember to permanently save your configurations to nonvolatile RAM (NVRAM) afterwards with the write memory or copy running-config startup-config.. This includes IP addresses... then you need to do very little editing of the saved configuration file. ! Enter the commands to provide a bare minimum of connectivity to ! your TFTP server here.1.qxd 2/25/05 10:07 AM Page 261 TFTP 261 ciscorouter#copy tftp: running-config Address or name of remote host []? 192. . Give your router the bare minimum configuration that allows it to ping your TFTP server (no access controls or routing protocols).file from 192.file.18Harrison_ch16.168.100 (via FastEthernet0/0): !!!!!! [OK .. Loading config.1. Reconnect the router to the networks on which it was originally connected. you can use TFTP to completely restore the configuration to the replacement device. Connect your router to the local network of the TFTP server. The procedure for restoring your configuration is simple: 1. 4. On rebooting. The commands you need are: ciscorouter> enable Password: ******** ciscorouter# write erase . 7. where you have to replace a router completely. 3. the router will copy the startup configuration stored in NVRAM into a clean running configuration environment.1.168. 6.100 Source filename []? config.100/config. .26521/52224 bytes] 26521 bytes copied in 1.

Making mistakes with TFTP can be fairly easy to do. CONCLUSION TELNET and TFTP are important applications in an overall systems administration strategy.. and you may want to consider automating the process by using a helper application.18Harrison_ch16. ciscorouter# copy tftp:file-name startup-config ciscorouter# reload Please be aware that the write erase command erases your NVRAM startup configuration. and xinetd Chapter 16 . making eavesdropping easier and more lucrative for the hacker. TELNET is a greater security risk as the connections are longer. TFTP sessions don’t even need a username and password.” .. such as expect. I’d suggest that you use an encrypted TELNET replacement whenever possible. “Secure Remote Logins and File Copying. TFTP. will be covered in Chapter 17. SSH. and the TFTP server process overwrites any existing file beneath its root directory in keeping with the instructions of the network engineer.qxd 2/25/05 10:07 AM Page 262 262 TELNET. always use it with great care. They both have the shortcoming of not encrypting their data and therefore need to be used on secured networks for improved security. One such product. and valuable usernames and passwords are exchanged.

and Secure Shell (SSH) is often used as a stealthy alternative to TELNET.qxd 2/25/05 10:08 AM Page 263 C H A P T E R 17 Secure Remote Logins and File Copying IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ A Quick Introduction to SSH Encryption Starting OpenSSH Testing the Status of SSH The /etc/ssh/sshd_config File Using SSH to Log Into a Remote Machine What to Expect with Your First Login Deactivating TELNET After Installing SSH Executing Remote Commands on Demand with SSH SCP: A Secure Alternative to FTP SFTP: Another Secure Alternative to FTP Using SSH and SCP Without a Password Conclusion One of the most popular file transfer and remote login Linux applications is OpenSSH. SSH and SCP clients are available for most operating systems including Windows. A QUICK INTRODUCTION TO SSH ENCRYPTION Data encryption is accomplished by using special mathematical equations to scramble the bits in a data stream to make it unreadable to anyone who does 263 .19Harrison_ch17. which provides a number of ways to create encrypted remote terminal and file transfer connections between clients and servers. The OpenSSH Secure Copy (SCP) and Secure FTP (SFTP) programs are secure replacements for FTP. OpenSSH isn’t limited to Linux.

The ~/ notation is a universally accepted way of referring to your home directory and is recognized by all Linux commands. Always investigate changes to be safe. reinstalling the operating system or upgrading the SSH application might regenerate the keys.ssh directory and need to be specially generated. This creates a situation in which the computers at each end of the SSH connection have each other’s public keys and are able to decrypt the data sent from the other end of the encrypted link or “tunnel”. All the public keys that an SSH client’s Linux user encounters are stored in a file named ~/. These files are also stored in your ~/. and authorized_keys stores all the authorized public keys from remote hosts that may log into your account without the need for passwords (more on this later). If there is a match then the login is authorized.qxd 2/25/05 10:08 AM Page 264 264 Secure Remote Logins and File Copying Chapter 17 not have access to the corresponding decryption equation. Linux also uses other key files at the user level to provide the capability of password-less logins and file copying to remote servers using SSH and SCP. Data encryption helps to prevent unauthorized users from having access to the data. The SSH client’s public key is uploaded to the server at the same time. SSH uses the concept of randomly generated private and public keys to do its encryption. but you have the option of regenerating them should they become compromised. as are all files and directories whose names begin with a period. Of course. If a key and IP address no longer match. The process is usually made even harder through the use of an encryption key that is used to modify the way the equations do the scrambling. The ls -a command lists all normal and hidden files in a directory.ssh/known_hosts along with the IP address that provided it.pub files are your private and public keys respectively. When you log into an SSH server. and then the client automatically sends its public key which the server uses to match against a predefined list in the user’s directory. Your server’s own public and private SSH keys are stored in the /etc/ssh/ directory. you are prompted as to whether you want to accept the download of the server’s public key before you can proceed. The keys are usually created only once. key changes can be caused by someone trying some sort of cyber attack as well. the SSH connection is established. A successful exchange of encrypted data requires the receiver to have a copy of the sender’s public key beforehand. The id_dsa and id_dsa. . For example. then SSH knows that something is wrong. Note The . In this case. Here’s how it’s done with SSH.19Harrison_ch17. You can recover the original data only if you have access to this key and the corresponding programs.ssh directory is a hidden directory.

Because SSH and SCP are part of the same application. stop. [root@bigboy tmp]# chkconfig sshd on sshd You can also start. You can configure SSH to start at boot by using the chkconfig command.1 #ListenAddress 0. TESTING THE STATUS OF SSH You can test whether the SSH process is running with: [root@bigboy tmp]# pgrep sshd You should get a response of plain old process ID numbers.qxd 2/25/05 10:08 AM Page 265 The /etc/ssh/sshd_config File 265 STARTING OPENSSH OpenSSH is installed by default during Linux installations. and restart SSH after booting by running the initialization script.19Harrison_ch17.0. they share the same configuration file and are governed by the same /etc/init. Take a look at a snippet from configuration: # The strategy used for options in the default sshd_config shipped with # OpenSSH is to specify options with their default value where # possible. Uncommented options change a # default value. THE /ETC/SSH/SSHD_CONFIG FILE The SSH configuration file is called /etc/ssh/sshd_config. By default SSH listens on all your NICs and uses TCP port 22. but leave them commented. [root@bigboy tmp]# service sshd start [root@bigboy tmp]# service sshd stop [root@bigboy tmp]# service sshd restart Remember to restart the SSH process every time you make a change to the configuration files for the changes to take effect on the running process.d/sshd startup script.0 #ListenAddress :: . #Port 22 #Protocol 2.0.

19Harrison_ch17. then you can change port 22 to a location that won’t interfere with other applications on your system.qxd 2/25/05 10:08 AM Page 266 266 Secure Remote Logins and File Copying Chapter 17 SSH Versions 1 and 2 The original encryption scheme of SSH was adequate for its time but was eventually found to have a number of limitations.0. Change the Port line in /etc/ssh/sshd_config to mention 435 and remove the # at the beginning of the line.1. such as port 435. pick another port and try again: Port 435 3. . Restart SSH: [root@bigboy tmp]# service sshd restart 4. using grep to filter out everything that doesn’t have the string “435”: [root@bigboy root]# netstat -an | grep 435 [root@bigboy root]# 2. Always force your systems to operate exclusively with version 2 by setting the protocol statement in the /etc/ssh/sshd_config file to 2. No response allows us to proceed. If port 435 is being used.0.100:435 0. Remember to restart SSH to make this take effect. What you need to do is: 1.0:* [root@bigboy root]# LISTEN Next. This is a rudimentary precaution only. # # File: /etc/ssh/sshd_config # Protocol 2 Change the TCP Port on Which SSH Listens If you are afraid of people trying to hack in on a well known TCP port. The answer to these was version 2. because good network scanning programs can detect SSH running on alternative ports.168. you’ll discover how to actually log into systems using SSH. Check to ensure SSH is running on the new port: [root@bigboy root]# netstat -an | grep 435 tcp 0 0 192. Use the netstat command to make sure your system isn’t listening on port 435.

19Harrison_ch17.qxd

2/25/05

10:08 AM

Page 267

What to Expect with Your First Login

267

USING SSH TO LOG INTO A REMOTE MACHINE
Using SSH is similar to TELNET. To log in from another Linux box use the ssh command with a -l to specify the username you wish to log in as. If you leave out the -l, your username will not change. Here are some examples for a server named Smallfry in your /etc/hosts file. You can also use the username@remote_server format as an alternative. If you are user root and you want to log into Smallfry as yourself, use the command:
[root@bigboy tmp]# ssh smallfry

User root can also log into Smallfry as user peter via the default port 22:
[root@bigboy tmp]# ssh -l peter smallfry

or via port 435 using the username@remote_server alternative login format:
[root@bigboy tmp]# ssh -p 435 peter@smallfry

WHAT TO EXPECT WITH YOUR FIRST LOGIN
The first time you log in, you get a warning message saying that the remote host doesn’t know about your machine and prompting you to store a copy of the remote host’s SSH identification keys on your local machine. It will look something like this:
[root@bigboy tmp]# ssh smallfry The authenticity of host ‘smallfry (smallfry)’ can’t be established. RSA key fingerprint is 5d:d2:f5:21:fa:07:64:0d:63:1b:3b:ee:a6:58:58:bb. Are you sure you want to continue connecting (yes/no)? yes Warning: Permanently added ‘smallfry’ (RSA) to the list of known hosts. root@smallfry’s password: Last login: Thu Nov 14 10:18:45 2002 from 192.168.1.98 No mail. [root@smallfry tmp]#

The key is stored in your ˜/.ssh/known_hosts file and you should never be prompted for this again.

19Harrison_ch17.qxd

2/25/05

10:08 AM

Page 268

268

Secure Remote Logins and File Copying

Chapter 17

SSH Failures Due to Linux Reinstallations If Linux or SSH is reinstalled on the remote server, then the keys are regenerated and your SSH client will detect that this new key doesn’t match the saved value in the known_hosts file. The SSH client will fail, erring on the side of caution to alert you to the possibility of a form of hacking attack:
[root@bigboy tmp]# ssh 192.168.1.102 @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ @ WARNING: REMOTE HOST IDENTIFICATION HAS CHANGED! @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ IT IS POSSIBLE THAT SOMEONE IS DOING SOMETHING NASTY! Someone could be eavesdropping on you right now (man-in-the-middle attack)! It is also possible that the RSA host key has just been changed. The fingerprint for the RSA key sent by the remote host is 5d:d2:f5:21:fa:07:64:0d:63:1b:3b:ee:a6:58:58:bb. Please contact your system administrator. Add correct host key in /root/.ssh/known_hosts to get rid of this message. Offending key in /root/.ssh/known_hosts:2 RSA host key for 192.168.1.102 has changed and you have requested strict checking. Host key verification failed. [root@bigboy tmp]#

If you are confident that the error is due to a reinstallation, then edit your ~/.ssh/known_hosts text file, removing the entry for the offending remote server. When you try connecting via SSH again, you’ll be prompted to add the new key to your ~/.ssh/known_hosts file and the login session should proceed as normal after that.

DEACTIVATING TELNET AFTER INSTALLING SSH
You should always consider SSH over TELNET, because of the inherent data encryption features of SSH and the current widespread availability of SSH clients for both Linux and Windows. By default, the TELNET server isn’t installed with Fedora Linux. If you do decide to deactivate an active TELNET server on Fedora, then use the chkconfig command as detailed in Chapter 16, “TELNET, TFTP, and xinetd.”
[root@bigboy tmp]# chkconfig telnet off

19Harrison_ch17.qxd

2/25/05

2:43 PM

Page 269

SCP: A Secure Alternative to FTP

269

EXECUTING REMOTE COMMANDS

ON

DEMAND WITH SSH

A nice feature of SSH is that it is capable of logging in and executing single commands on a remote system. You just have to place the remote command, enclosed in quotes, at the end of the ssh command of the local server. In the example below, a user on server Smallfry who needs to know the version of the kernel running on server Bigboy (192.168.1.100) remotely runs the uname -a command. The command returns the version of 2.6.8-1.521 and the server’s name, Bigboy:
[root@smallfry tmp]# ssh 192.168.1.100 “uname -a” root@192.168.1.100’s password: Linux bigboy 2.6.8-1.521 #1 Mon Aug 16 09:01:18 EDT 2004 i686 i686 i386 GNU/Linux [root@smallfry tmp]#

This feature can be very useful. You can combine it with password-free login, explained later in this chapter, to get the status of a remote server whenever you need it. More comprehensive monitoring may best be left to such purpose built programs as MRTG, which is covered in Chapter 22, “Monitoring Server Performance.”

SCP: A SECURE ALTERNATIVE TO FTP
From a networking perspective, FTP isn’t very secure, because usernames, passwords, and data are sent across the network unencrypted. More secure forms such as SFTP (Secure FTP) and SCP (Secure Copy) are available as a part of the OpenSSH package that is normally installed by default on Red Hat and Fedora Core. Remember, unlike FTP, SCP doesn’t support anonymous downloads like FTP. The Linux scp command for copying files has a format similar to that of the regular Linux cp command. The first argument is the source file and the second is the destination file. When copying to or from a remote server, SCP logs into the server to transfer the data and this therefore requires you to supply a remote server name, username, and password to successfully execute the command. The remote filename is therefore preceded by a prefix of the remote username and server name separated by an @ symbol. The remote filename or directory then follows separated by a colon. The format therefore looks like this:
username@servername:filename username@servername:directoryname

19Harrison_ch17.qxd

2/25/05

10:08 AM

Page 270

270

Secure Remote Logins and File Copying

Chapter 17

For example, file /etc/syslog.conf on a server with IP address 192.168.1.100 that needs to be retrieved as user peter would have the format peter@192.168.1.000:/etc/syslog.conf, the entire /etc directory would be
peter@192.168.1.000:/etc/.

Tip You can download an easy-to-use Windows SCP client called WinSCP from
http://winscp.vse.cz/eng/.

Copying Files to the Local Linux Box If you understand how scp represents remote filenames, you can start copying files fairly easily. For example, to copy file /tmp/software.rpm on the remote machine to the local directory /usr/rpm use the commands:
[root@bigboy tmp]# scp root@smallfry:/tmp/software.rpm /usr/rpm root@smallfry’s password: software.rpm 100% 1011 27.6KB/s 00:00 [root@bigboy tmp]#

To copy the file /tmp/software.rpm on the remote machine to the local directory /usr/rpm using TCP port 435, use the commands:
[root@bigboy tmp]# scp -P 435 root@smallfry:/tmp/software.rpm /usr/rpm root@smallfry’s password: software.rpm 100% 1011 27.6KB/s 00:00 [root@bigboy tmp]#

Copying Files to the Remote Linux Box Copying files to the local Linux server now becomes intuitive. For example, to copy file /etc/hosts on the local machine to directory /tmp on the remote server:
[root@bigboy tmp]# scp /etc/hosts root@192.168.1.103:/tmp root@192.168.1.103’s password: hosts 100% 1011 27.6KB/s 00:00 [root@bigboy tmp]#

To copy file /etc/hosts on the local machine to directory remote server via TCP port 435, use the commands:

/tmp

on the

[root@bigboy tmp]# scp -P 435 /etc/hosts root@192.168.1.103:/tmp hosts 100% 1011 27.6KB/s 00:00 [root@bigboy tmp]#

19Harrison_ch17.qxd

2/25/05

10:08 AM

Page 271

Using SSH and SCP without a Password

271

SFTP: ANOTHER SECURE ALTERNATIVE TO FTP
OpenSSH also has the SFTP program, which runs over an encrypted SSH session but whose commands mimic those of FTP. SFTP can be more convenient to use than SCP when you are uncertain of the locations of the files you want to copy, because it has the directory browsing abilities of FTP. Unlike FTP, SFTP doesn’t support anonymous logins. Here is a sample login sequence that logs in, gets help on the available commands, and downloads a file to the local server:
[root@bigboy tmp]# sftp 192.168.1.200 Connecting to 192.168.1.200... root@192.168.1.200’s password: sftp> help Available commands: cd path Change remote directory to ‘path’ lcd path Change local directory to ‘path’ chgrp grp path Change group of file ‘path’ to ‘grp’ chmod mode path Change permissions of file ‘path’ to ‘mode’ ... ... sftp> ls .. .. anaconda-ks.cfg install.log install.log.syslog .. .. sftp> get install.log install.log 100% 17KB 39.4KB/s 00:00 sftp> exit [root@bigboy tmp]#

USING SSH

AND

SCP WITHOUT

A

PASSWORD

From time to time you may want to write scripts that will allow you to copy files to a server, or login, without being prompted for passwords. This can make them simpler to write and also prevents you from having to embed the password in your code. SCP has a feature that allows you to do this. You no longer have to worry about prying eyes seeing your passwords nor worry about your script breaking when someone changes the password. You can configure SSH to do this by generating and installing data transfer encryption keys that are tied to the IP addresses of the two servers. The servers then use these pre-installed keys to authenticate one another for each file transfer. As you may expect, this feature

19Harrison_ch17.qxd

2/25/05

2:43 PM

Page 272

272

Secure Remote Logins and File Copying

Chapter 17

doesn’t work well with computers with IP addresses that periodically change, such as those obtained via DHCP. There are some security risks though. The feature is automatically applied to SSH as well. Someone could use your account to log into the target server by entering the username alone. It is therefore best to implement this using unprivileged accounts on both the source and target servers. The example that follows enables this feature in one direction (from server Bigboy to server Smallfry) and only uses the unprivileged account called filecopy. Configuration: Client Side Here are the steps you need to do on the computer that acts as the SSH client: 1. Generate your SSH encryption key pair for the filecopy account. Press the Enter key each time you are prompted for a password to be associated with the keys. (Do not enter a password.)
[filecopy@bigboy filecopy]# ssh-keygen -t dsa Generating public/private dsa key pair. Enter file in which to save the key (/filecopy/.ssh/id_dsa): Enter passphrase (empty for no passphrase): Enter same passphrase again: Your identification has been saved in /filecopy/.ssh/id_dsa. Your public key has been saved in /filecopy/.ssh/id_dsa.pub. The key fingerprint is: 1e:73:59:96:25:93:3f:8b:50:39:81:9e:e3:4a:a8:aa filecopy@bigboy [filecopy@bigboy filecopy]#

2. These keyfiles are stored in the .ssh subdirectory of your home directory. View the contents of that directory. The file named id_dsa is your private key, and id_dsa.pub is the public key that you will be sharing with your target server. Versions other than Red Hat/Fedora may use different filenames, use the SSH man pages to verify this.
[filecopy@bigboy filecopy]# cd ~/.ssh [filecopy@bigboy filecopy]# ls id_dsa id_dsa.pub known_hosts [filecopy@bigboy .ssh]#

3. Copy only the public key to the home directory of the account to which you will be sending the file.
[filecopy@bigboy .ssh]# scp id_dsa.pub \ filecopy@smallfry:public-key.tmp

Now, on to the server side of the operation.

19Harrison_ch17.qxd

2/25/05

2:44 PM

Page 273

Conclusion

273

Configuration: Server Side Here are the steps you need to do on the computer that will act as the SSH server. 1. Log into Smallfry as user filecopy. Create an .ssh subdirectory in your home directory and then go to it with cd.
[filecopy@smallfry public-key.tmp [filecopy@smallfry [filecopy@smallfry [filecopy@smallfry filecopy]# ls filecopy]# mkdir .ssh filecopy]# chmod 700 .ssh filecopy]# cd .ssh

2. Append the public-key.tmp file to the end of the authorized_keys file using the >> append redirector with the cat command. The authorized_keys file contains a listing of all the public keys from machines that are allowed to connect to your Smallfry account without a password. Versions other than Red Hat/Fedora may use different filenames, use the SSH man pages to verify this.
[filecopy@smallfry .ssh]# cat ~/public-key.tmp >> authorized_keys [filecopy@smallfry .ssh]# rm ~/public-key.tmp

From now on you can use ssh and scp as user filecopy from server Bigboy to Smallfry without being prompted for a password.

CONCLUSION
Most Linux security books strongly recommend using SSH and SCP over TELNET and FTP because of their encryption capabilities. Despite this, there is still a place for FTP in the world thanks to its convenience in providing simple global access to files and TELNET, which is much easier to implement in price-sensitive network appliances than SSH. Consider all options when choosing your file transfer and remote login programs and select improved security whenever possible as the long-term benefits eventually outweigh the additional cost over time.

19Harrison_ch17.qxd

2/25/05

10:08 AM

Page 274

20Harrison_ch18.qxd

2/25/05

10:09 AM

Page 275

C H A P T E R

18

Configuring DNS

IN THIS CHAPTER
☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Introduction to DNS Downloading and Installing the BIND Packages Starting BIND The /etc/resolv.conf File Configuring a Caching Name Server Important File Locations Configuring a Regular Name Server Troubleshooting BIND Migrating Your Web Site In-House DHCP Considerations for DNS Simple DNS Security Conclusion

Domain Name System (DNS) converts the name of a Web site (www.linuxhometo an IP address (65.115.71.34). This step is important, because the IP address of a Web site’s server, not the Web site’s name, is used in routing traffic over the Internet. This chapter will explain how to configure your own DNS server to help guide Web surfers to your site.
networking.com)

INTRODUCTION TO DNS
Before you dig too deep in DNS, you need to understand a few foundation concepts on which the rest of the chapter will be built.

275

20Harrison_ch18.qxd

2/25/05

10:09 AM

Page 276

276

Configuring DNS

Chapter 18

DNS Domains Everyone in the world has a first name and a last, or family, name. The same thing is true in the DNS world: A family of Web sites can be loosely described as a domain. For example, the domain linuxhomenetworking.com has a number of children, such as www.linuxhomenetworking.com and mail.linuxhomenetworking.com for the Web and mail servers, respectively. BIND BIND is an acronym for the Berkeley Internet Name Domain Project, which is a group that maintains the DNS-related software suite that runs under Linux. The most well-known program in BIND is named, the daemon that responds to DNS queries from remote machines. DNS Clients A DNS client doesn’t store DNS information; it must always refer to a DNS server to get it. The only DNS configuration file for a DNS client is the /etc/resolv.conf file, which defines the IP address of the DNS server it should use. You shouldn’t need to configure any other files. You’ll become well acquainted with the /etc/resolv.conf file soon. Authoritative DNS Servers Authoritative servers provide the definitive information for your DNS domain, such as the names of servers and Web sites in it. They are the last word in information related to your domain. How DNS Servers Find Your Site Information There are 13 root authoritative DNS servers (super duper authorities) that all DNS servers query first. These root servers know all the authoritative DNS servers for all the main domains—.com, .net, and the rest. This layer of servers keep track of all the DNS servers that Web site systems administrators have assigned for their sub domains. For example, when you register your domain my-web-site.org, you are actually inserting a record on the .com DNS servers that point to the authoritative DNS servers you assigned for your domain. (More on how to register your site later.) When to Use a DNS Caching Name Server Most servers don’t ask authoritative servers for DNS directly, they usually ask a caching DNS server to do it on their behalf. The caching DNS servers then store (or cache) the most frequently requested information to reduce the lookup overhead of subsequent queries.

20Harrison_ch18.qxd

2/25/05

10:09 AM

Page 277

Introduction to DNS

277

If you want to advertise your Web site www.my-web-site.org to the rest of the world, then a regular DNS server is what you require. Setting up a caching DNS server is fairly straightforward and works whether or not your ISP provides you with a static or dynamic IP address. After you set up your caching DNS server, you must configure each of your home network PCs to use it as their DNS server. If your home PCs get their IP addresses using DHCP, then you have to configure your DHCP server to make it aware of the IP address of your new DNS server, so that the DHCP server can advertise the DNS server to its PC clients. Off-the-shelf router/firewall appliances used in most home networks usually can act as both the caching DNS and DHCP server, rendering a separate DNS server is unnecessary. You can find the configuration steps for a Linux DHCP server in Chapter 8, “Configuring the DHCP Server.” When to Use a Static DNS Server If your ISP provides you with a fixed or static IP address, and you want to host your own Web site, then a regular authoritative DNS server would be the way to go. A caching DNS name server is used as a reference only, regular name servers are used as the authoritative source of information for your Web site’s domain. Note Regular name servers are also caching name servers by default.

When To Use A Dynamic DNS Server If your ISP provides your router/firewall with its IP address using DHCP, then you must consider dynamic DNS covered in Chapter 19, “Dynamic DNS.” For now, I’m assuming that you are using static IP addresses. How to Get Your Own Domain Whether or not you use static or dynamic DNS, you need to register a domain. Dynamic DNS providers frequently offer you a subdomain of their own site, such as my-web-site.dnsprovider.com, in which you register your domain on their site. If you choose to create your very own domain, such as my-web-site.org, you have to register with a company specializing in static DNS registration and then point your registration record to the intended authoritative DNS for your domain. Popular domain registrars include VeriSign, Register Free, and Yahoo. If you want to use a dynamic DNS provider for your own domain, then you have to point your registration record to the DNS servers of your dynamic DNS provider. (More details on domain registration are coming later in the chapter.)

20Harrison_ch18.qxd

2/25/05

10:09 AM

Page 278

278

Configuring DNS

Chapter 18

Basic DNS Testing of DNS Resolution As you know, DNS resolution maps a fully qualified domain name (FQDN), such as www.linuxhomenetworking.com, to an IP address. This is also known as a forward lookup. The reverse is also true: By performing a reverse lookup, DNS can determining the fully qualified domain name associated with an IP address. Many different Web sites can map to a single IP address, but the reverse isn’t true; an IP address can map to only one FQDN. This means that forward and reverse entries frequently don’t match. The reverse DNS entries are usually the responsibility of the ISP hosting your site, so it is quite common for the reverse lookup to resolve to the ISP’s domain. This isn’t an important factor for most small sites, but some e-commerce applications require matching entries to operate correctly. You may have to ask your ISP to make a custom DNS change to correct this. There are a number of commands you can use to do these lookups. Linux uses the host command, for example, but Windows uses nslookup.
The host Command

The host command accepts arguments that are either the fully qualified domain name or the IP address of the server when providing results. To perform a forward lookup, use the syntax:
[root@bigboy tmp]# host www.linuxhomenetworking.com www.linuxhomenetworking.com has address 65.115.71.34 [root@bigboy tmp]#

To perform a reverse lookup:
[root@bigboy tmp]# host 65.115.71.34 34.71.115.65.in-addr.arpa domain name pointer 65-115-71-34.my-ispprovider.net. [root@bigboy tmp]#

As you can see, the forward and reverse entries don’t match. The reverse entry matches the entry of the ISP.
The nslookup Command

The nslookup command provides the same results on Windows PCs. To perform forward lookup, use:
C:\> nslookup www.linuxhomenetworking.com Server: 192-168-1-200.my-web-site.org Address: 192.168.1.200 Non-authoritative answer: Name: www.linuxhomenetworking.com Address: 65.115.71.34 C:\>

20Harrison_ch18.qxd

2/25/05

10:09 AM

Page 279

The /etc/resolv.conf File

279

To perform a reverse lookup:
C:\> nslookup 65.115.71.34 Server: 192-168-1-200.my-web-site.org Address: 192.168.1.200 Name: 65-115-71-34.my-isp-provider.net Address: 65.115.71.34 C:\>

DOWNLOADING

AND INSTALLING THE

BIND PACKAGES

Most Red Hat and Fedora Linux software products are available in the RPM format. When searching for the file, remember that the BIND RPM’s filename usually starts with the word “bind” followed by a version number, as in bind9.2.2.P3-9.i386.rpm. (For more details on downloading RPMs, see Chapter 6, “Installing RPM Software.”)

STARTING BIND
You can use the chkconfig command to get BIND configured to start at boot:
[root@bigboy tmp]# chkconfig named on

To start, stop, and restart BIND after booting, use:
[root@bigboy tmp]# service named start [root@bigboy tmp]# service named stop [root@bigboy tmp]# service named restart

Remember to restart the BIND process every time you make a change to the configuration file for the changes to take effect on the running process.

THE /ETC/RESOLV.CONF FILE
DNS clients (servers not running BIND) use the /etc/resolv.conf file to determine both the location of their DNS server and the domains to which they belong. The file generally has two columns; the first contains a keyword, and the second contains the desired values separated by commas. See Table 18.1 for a list of keywords.

1.org and my-web-business.org.102. If the server is bigboy.168.1.org. but it also is a member of domains another-website.100 nameserver 192.1.qxd 2/25/05 10:09 AM Page 280 280 Configuring DNS Chapter 18 Table 18. your old entry of: nameserver 192.org another-web-site.102 The first domain listed after the search directive must be the home domain of your network. If there is more than one name server. then the entry would just be my-web-site. enabling you to refer to servers in the same domain by only their server name without having to specify the domain.168.org.0.20Harrison_ch18.1.conf. So. This is a handy time-saving feature. provide DNS name resolution: search my-web-site. my-web-site.org nameserver 192. There should be only one entry per nameserver keyword. which should be searched for shorthand references to other servers. If you refer to another server just by its name without the domain added on. Two name servers. CONFIGURING A CACHING NAME SERVER The Red Hat/Fedora default installation of BIND is configured to convert your Linux box into a caching name server.org. The domains in this list must be separated by spaces.168.1. use multiple nameserver lines. DNS on your client appends the server name to each domain in this list and does a DNS lookup on each to get the remote servers’ IP addresses.1 . in this case my-web-site.100 and 192.0.1 would be replaced by a new entry of: # nameserver 192.1. Placing a domain and search entry in the /etc/resolv. nameserver domain search Take a look at a sample configuration in which the client server’s main domain is my-web-site.conf is redundant.1 Keyword Keywords in /etc/resolv. therefore.168.168. you’ll have to comment out the reference to your previous DNS server (most likely your router) with a # or make it point to the server itself using the universal localhost IP address of 127. The local domain name to be used by default. The only file you have to edit is /etc/resolv.1.org. 192.org my-web-business.conf Value IP address of your DNS name server.168.

it is important to understand exactly where the files are located. You can determine whether you have the chroot add-on RPM by using this command. . and the chroot BIND add-on RPM installs its own versions in their chroot locations. Table 18.3-13 [root@bigboy tmp]# There can be confusion with the locations: Regular BIND installs its files in the normal locations.* /var/named/chroot/etc/ Before you go to the next step of configuring a regular name server. but also to limit the files named can see. IMPORTANT FILE LOCATIONS Red Hat/Fedora BIND normally runs as the named process owned by the unprivileged named user.0. When installed.qxd 2/25/05 10:09 AM Page 281 Important File Locations 281 or nameserver 127. the hacker’s access to the rest of your system is isolated to the files under the chroot directory and nothing else. Unfortunately.conf /var/named/chroot/etc/ [root@bigboy tmp]# cp -f /etc/rndc. Sometimes BIND is also installed using Linux’s chroot feature to not only run named as user named.2.1 The next step is to make all the other machines on your network point to the caching DNS server as their primary DNS server.20Harrison_ch18. The advantage of the chroot feature is that if a hacker enters your system via a BIND exploit. the chroot versions of some of the files are empty. Therefore. named files normally found in the /etc directory are found in /var/named/chroot/etc directory instead. Before starting Fedora BIND. This type of security is also known as a chroot jail. named is fooled into thinking that the directory /var/named/chroot is actually the root or / directory. copy the configuration files to their chroot locations: [root@bigboy tmp]# cp -f /etc/named. and those you’d expect to find in /var/named are actually located in /var/named/chroot/var/named.2 provides a map. which returns the name of the RPM: [root@bigboy tmp]# rpm -q bind-chroot bind-chroot-9.0.

conf file to reference localhost only. you’ll set up the forward zone for www.255.24 with a subnet mask of 255.conf rndc.conf /var/named/chroot/etc /var/named/chroot/var/named /etc /var/named zone files Note Fedora Core installs BIND chroot by default.158.248 (/29). CONFIGURING A REGULAR NAME SERVER For the purposes of this tutorial.255.key.zone.conf file. Use the code: .1 Configuring named.conf You’ll have to make your DNS server refer to itself for all DNS queries by configuring the /etc/resolv.my-web-site. Red Hat 9 and earlier versions don’t.conf file contains the main DNS configuration and tells BIND where to find the configuration files for each domain you own.conf The named.zone should be located in the default directory of /var/named/chroot/var/named in a chroot or in /var/named in a regular one.qxd 2/25/05 2:45 PM Page 282 282 Configuring DNS Chapter 18 Table 18. This file usually has two zone areas: ☞ Forward zone file definitions list files to map domains to IP addresses ☞ Reverse zone file definitions list files to map IP addresses to domains In this example.253. nameserver 127.0. and.0. rndc. although not explicitly stated. the file my-site. assume your ISP assigned you the subnet 97.2 File Differences in Fedora and Red Hat DNS File Locations Purpose Tells the names of the zone files to be used for each of your Web site domains Named authentication Link all the IP addresses in your domain to their corresponding servers BIND chroot Location /var/named/chroot/etc Regular BIND Location /etc named.20Harrison_ch18. The zone file is named my-site.org by placing entries at the bottom of the named. Configuring resolv.

This isn’t important for the Windows clients on your network. file “another-site.arpa domain. you have to format entries to handle the reverse lookups for your IP addresses. The forward domain lookup process for mysite. Note The allow-query directive defines the networks that are allowed to query your DNS server for information on any zone. you could modify the directive to: allow-query { 192.com” { type master. file “my-site.zone”. but the noticeable difference is that for forward lookups the scan is from right to left.168.com using a zone file named another-site.zone for the 192.1. but you will have to create reverse zone entries for your SOHO/home environment using the 192. }. }. and for reverse lookups the scan is from left to right. }. For example. you can insert additional entries in the named.zone”. your ISP handles the reverse zone entries for your public IP addresses. }.0/24. In addition. to which all IP addresses belong. This order is important to remember or else the configuration will fail.com scans the FQDN from right to left to get increasingly more specific information about the authoritative servers to use.org” { type master.zone: zone “another-site. The similarity in both methods is that increasingly specific information is sought.168.0/24 network: .0 network.qxd 2/25/05 10:09 AM Page 283 Configuring a Regular Name Server 283 zone “my-web-site.1. Next.168. allow-query { any. notify no. }. to limit queries to only your 192. allow-query { any. but some Linux applications require valid forward and reverse entries to operate correctly.conf uses a reverse zone file named 192-168-1. In most cases.conf file to reference other Web domains you host.0/24 address space. notify no. This reverse zone definition for named. Here is an example for another-site.1. is followed by the first three octets of the IP address in reverse order.168.1.conf file where the main in-addr.20Harrison_ch18. Reverse lookups operate similarly by scanning an IP address from left to right to get increasingly specific information about an address. This difference can be seen in the formatting of the zone statement for a reverse zone in /etc/named.

}.196. ☞ By default.qxd 2/25/05 10:09 AM Page 284 284 Configuring DNS Chapter 18 zone “1.in-addr.arpa” { type master. and CNAME) that govern different areas of BIND.zone”. A. If the TTL is set to three days. a W signifies weeks. In the absence of a suffix. notify no. forward lookups (A). and an H signifies hours. Configuring the Zone Files You need to keep a number of things in mind when configuring DNS zone files: ☞ In all zone files. . DNS Resource Records The rest of the records in a zone file are usually BIND resource records.20Harrison_ch18. Take a closer look at these entries in the zone file. ☞ Each zone file contains a variety of records (SOA. Time to Live Value The very first entry in the zone file is usually the zone’s time to live (TTL) value. The authoritative servers not only provide the DNS answer but also provide the information’s time to live.168. then caching servers use the original stored response for three days before making the query again: $TTL 3D BIND recognizes several suffixes for time-related values. which is the period for which it’s valid. They define the nature of the DNS information in your zone files that’s presented to querying DNS clients. They all have the general format: Name Class Type Data There are different types of records for mail (MX). aliases (CNAME) and overall zone definitions. Caching DNS servers cache the responses to their queries from authoritative DNS servers. you can place a comment at the end of any line by inserting a semi-colon character then typing in the text of your comment. reverse lookups (PTR). your zone files are located in the directory /var/named or /var/named/chroot/var/named. A D signifies days. BIND assumes the value is in seconds. The purpose of a TTL is to reduce the number of DNS queries the authoritative DNS server has to answer. NS. file “192-168-1. MX.

The regular @ in the e-mail address must be replaced with a period instead. Fully qualified name of your primary name server.conf file for that particular database file. you can insert new line characters between the fields as long as you insert parenthesis at the beginning and end of the insertion to alert BIND that part of the record will straddle multiple lines. Here is an example: @ IN SOA ns1.3 explains what each field in the record means. Other classes exist for non-Internet protocols and functions but are very rarely used. It has the format: Name Class Type Name-Server Email-Address Serial-No Refresh Retry Expiry Minimum-TTL The record can be long. which contains general administrative and control information about the domain.qxd 2/25/05 10:09 AM Page 285 Configuring a Regular Name Server 285 Start of Authority (SOA). Must be followed by a period. ( serial # refresh retry 1W 1D ) . this is an SOA resource record. the name is also subject to interpretation based on this factor. The e-mail address of the name server administrator.my-web-site. Similarly. which I’ll cover later. 1H .my-web-site. minimum Table 18. hostmaster. There are a number of different DNS classes. The e-mail address must also be followed by a period. Table 18. The data portion is formatted according to the record type and may consist of several values separated by spaces. 4H . Other types of records exist. Home/SOHO will be limited to the IN or Internet class used when defining IP address mapping information for BIND. The @ sign is a shorthand reference to the current origin (zone) in the /etc/named. You can also add comments to the end of each new line separated by a semicolon when you do this. For the sake of formatting. 2004100801 . expiry . You can use the format YYYYMMDD with single digit incremented number tagged to the end to provide an incremental value that provides some editing information.3 Field Name Class The SOA Record Format Description The root name of the zone. The SOA Record The first resource record is the SOA record. A serial number for the current configuration. The type of DNS resource record.20Harrison_ch18.org. In the example.org. and will sometimes wrap around on your screen. Type Name-server Email-address Serial-no continues .

This value defines the caching duration your DNS includes in this response. Your DNS server will respond with a no domain or NXDOMAIN response that the remote client caches. MX. Minimum-TTL So in the example. Table 18. and minimum values of 4 hours.org with a contact e-mail address of hostmaster@my-web-site. The slave’s retry interval to connect the master in the event of a connection failure. PTR. A. You should also be aware that IN is the default Class. Slaves aren’t usually used in home/SOHO environments. the Class field will always be IN or Internet. For most home/SOHO scenarios.4 outlines the way they are laid out. Name of a server in the domain Server name alias Last octet of server’s IP address Class Field2 IN IN Type Field NS MX Data Field IP address or CNAME of the name server Mail server DNS name A CNAME PTR IN IN IN A CNAME PTR IP address of server “A” record name for the server Fully qualified server name 1 2 If the search key to a DNS resource record is blank. Table 18. the NS. and CNAME records each occupy a single line with a very similar general format. and 1 day. expiry. Future references will be directed towards the root servers.my-web-site. usually the same as the domain of the zone file itself. A. There are times when remote clients will make queries for subdomains that don’t exist. Slaves aren’t usually used in home/SOHO environments. Slaves aren’t usually used in home/SOHO environments. and BIND assumes a record is of this type unless otherwise stated.3 Field Refresh Retry Expiry continued Description Tells the slave DNS server how often it should check the master DNS server. MX. Total amount of time a slave should retry to contact the master before expiring the data it contains. NS. For the example.20Harrison_ch18. and CNAME Record Formats Field Descriptions Name Field NS MX Usually blank1 Domain to be used for mail. PTR. . PTR.org. and CNAME Records Like the SOA record.qxd 2/25/05 10:09 AM Page 286 286 Configuring DNS Chapter 18 Table 18.4 Record Type NS. retry. 1 week. it reuses the search key from the previous record. respectively. 1 hour. the primary name server is defined as ns1. A. MX. this is the SOA @ sign. The serial number is 2004100801 with refresh.

26 bigboy bigboy bigboy Notice that in this example: ☞ Server ns1.myweb-site.0.my-web-site.my-web-site.org: .org. BIND attaches the my-web-site. Inet Address of nameserver my-web-site.0. NS. ☞ The MX record for my-web-site.253.qxd 2/25/05 10:09 AM Page 287 Configuring a Regular Name Server 287 If you don’t put a period at the end of a host name in a SOA. hostmaster. localhost bigboy mail ns1 www MX 10 mail .org . retry. therefore remote DNS caching servers will store learned DNS information from your zone for three days before flushing it out of their caches.20Harrison_ch18. In corporate environments there may be a separate name server for this purpose.my-web-site. The full zone file .mysite.my-website.org. serial# 3600 .158.com. BIND assumes an A record with www refers to www. Zone file for my-web-site.1 97. seconds .org points to the server named mail. Sample Forward Zone File Now that you know the key elements of a zone file. or CNAME record. This may be acceptable in most cases.org. . Primary name servers are more commonly called ns1 and secondary name servers ns2.org.org. ( 200211152 . minimum. seconds 3600 . refresh.org. NS www . So. A. expire. $TTL 3D @ IN SOA ns1. seconds 3600 ) . and you will find your mail server accepting mail only for the domain my-web-site. . ☞ The minimum TTL value ($TTL) is three days. but if you forget to put the period after the domain in the MX record for my-web-site.org at the end. .org is the name server for my-web-site. seconds 3600 . it’s time to examine a working example for the domain my-web-site.org .org. Primary Mail Exchanger A A CNAME CNAME CNAME 127. BIND will automatically tack on the zone file’s domain name to the name of the host.

168. which is used in network-based file access. This is an example of a zone file for the 192. seconds . also requires valid reverse lookup capabilities. .158. mail server. seconds retry.org. then you’d have an A record entry for each: www mail ns A A A 97. Sample Reverse Zone File Now you need to make sure that you can do a host query on all your home network’s PCs and get their correct IP addresses.253. and use the serial number to determine whether the data on the master has been updated. NS . Zone file for 192. and Web server being the same machine. ( www.158.x network.1. .20Harrison_ch18.org. .100 points to the name bigboy.my-web-site.my-webserial number refresh.134 97. www .qxd 2/25/05 10:09 AM Page 288 288 Configuring DNS Chapter 18 ☞ ns1 and mail are actually CNAMEs or aliases for the Web server www.26 97. Notice how the main difference between forward and reverse zone files is that the reverse zone file only has PTR and NS records.168. 200303301 8H 2H 4W 1D ) .253. If they were all different machines. seconds minimum. .253. so the IP address 192. All the entries in the first column refer to the last octet of the IP address for the network. could cause your slaves to have outdated information. because sendmail typically relays mail only from hosts whose IP addresses resolve correctly in DNS.1.org.zone . NFS. seconds expire. . Also the PTR records cannot have CNAME aliases.x . Nameserver Address .158u. 100 PTR bigboy. hostmaster. the slave DNS servers periodically poll the master server for updated zone file information.my-web-site.org. Filename: 192-168-1. When DNS is set up in a redundant configuration. .168. Failing to increment the serial number.1.my-website. $TTL 3D @ IN SOA site. even though the contents of the zone file have been modified.125 It is a required practice to increment your serial number whenever you edit your zone file. This is very important if you are running a mail server on your network. So here you have an example of the name server. .

1. This means it’s possible to use one set of zone files for queries from the Internet and another set for queries from your home network.X network.org.X. which are the addresses the DHCP server issues.253.253.20Harrison_ch18. ochorios.my-web-site. Here’s a summary of how it’s done: 1. This entry would be inserted in the internal section. The first section is called internal and lists the zone files to be used by your internal network. dhcp-192-168-1-32.26 address of the Web server. Place your zone statements in the /etc/named.org. What You Need to Know About NAT and DNS The previous examples assume that the queries will be coming from the Internet with the zone files returning information related to the external 97. SMTP mail relay wouldn’t work for PCs that get their IP addresses via DHCP if these lines weren’t included.my-web-site.36. dhcp-192-168-1-36.26 NAT IP address of your Web server.158. dhcp-192-168-1-35.org. For example.org.org.158. You may also want to create a reverse zone file for the public NAT IP addresses for your home network.158.conf file in one of two views sections. Most home DSL sites wouldn’t qualify. dhcp-192-168-1-33. . This /etc/named. You could also have a file called my-site-home. which Internet users would see. dhcp-192-168-1-34.168.158.zone for lookups related to the 97. Unfortunately.my-web-site.32 to 192.org.1.zone for lookups by home users on the 192. 192.1.168.my-web-site. The second view called external lists the zone files to used for Internet users. Creating the my-site-home.0 network.my-web-site.my-web-site.253. because NAT won’t work properly if a PC on your home network attempts to connect to the external 97.100. you could have a reference to a zone file called my-site. What do the PCs on your home network need to see? They need to see DNS references to the real IP address of the Web server. Don’t worry.168.qxd 2/25/05 10:09 AM Page 289 Configuring a Regular Name Server 289 103 102 105 32 33 34 35 36 PTR PTR PTR PTR PTR PTR PTR PTR smallfry.168.168.org. I included entries for addresses 192. BIND has a way around this called views.253.my-web-site.zone file and replace all references to 97.zone file is fairly easy: Copy it from the my-site.org.my-web-site.1. ISPs won’t usually delegate this ability for anyone with less than a Class C block of 256 IP addresses.conf entry would be inserted in the external section. The views feature allows you to force BIND to use predefined zone files for queries from certain subnets.X with references to 192.1. reggae.

0.0. }. }.168.0.0.0.0.0.0. . “trusted-subnet”.20Harrison_ch18.in-addr.0.0.127.0. To do this.0. Some built-in ACLs can save you time: ☞ ☞ ☞ Refers to the DNS server itself Refers to all the networks to which the DNS server is directly connected any: which is self explanatory localnets: localhost: Note You must place your localhost.0.0. allow-update { none.ip6. you must first define the internal and external networks with access control lists (ACLs) and then refer to these lists within their respective view section with the match-clients statement. file “localhost.0. file “named. }. Remember to increment your serial numbers! Here is a sample configuration snippet for the /etc/named. }.ca”.0.local”.ip6.0.127.” zone statements in the internal views section. zone “0.0. and period “.0.0. All the statements below were inserted after the options and controls sections in the file. file “named. 0.17. view “internal” { // What the home network will see match-clients { localnets. localhost.0.0. You must also tell the DNS server which addresses you feel are internal and external.0. allow-update { none. }.0/24. // IPv6 Support zone “0.0.conf file I use for my home network.arpa” IN { type master. zone “.0.arp a” IN { type master. file “named.0.0. }. }.local”. // ACL statement acl “trusted-subnet” { 192. }.arpa.0.0.in-addr.qxd 2/25/05 10:09 AM Page 290 290 Configuring DNS Chapter 18 2. zone “localhost” IN { type master.0.zone”. allow-update { none.” IN { type hint.0.0.

}.org” { type master. zone “another-site. zone “1. }.0” zone “0. }. }. }. file “my-site.arpa” IN { type master. }. allow-query { any. allow-query { any. allow-query { any. zone “my-web-site. }.broadcast”.arpa” IN { type master.arpa” IN { type master.qxd 2/25/05 10:09 AM Page 291 Configuring a Regular Name Server 291 }. }.com” { type master.192. file “my-site-home. // Prevents lookups for network addresses ending in “. }.20Harrison_ch18. // Prevents lookups for broadcast addresses ending in “.in-addr.zone”. allow-query { any.in-addr. }.zero”. zone “my-web-site. view “external” { // What the Internet will see match-clients { any. }.in-addr. notify no. recursion no. allow-update { none. notify no. }.zone”. }.com” { type master.zone”. file “named. file “another-site-home.zone”.255” zone “255.org” { type master.zone”. allow-update { none. }. }.168. file “named. file “another-site. file “192-168-1. }. }. notify no. . zone “another-site. notify no. allow-update { none.

but be sure to increment your configuration file serial number before doing this: [root@bigboy tmp]# service named restart Take a look at the end of your are no errors.local my-site. .1 -rw-r--r-. The network won’t need the trusted-subnet section in the match-clients line either.ca named.1 -rw-r--r-.1 -rw-r--r-.qxd 2/25/05 10:09 AM Page 292 292 Configuring DNS Chapter 18 In this example.1 -rw-r--r-. /var/log/messages file to make sure there Make Sure Your /etc/hosts File is Correctly Updated Chapter 3.1 -rw-r--r-.” explains how to correctly configure your /etc/hosts file.168.zone named.0).1 [root@bigboy tmp]# cd /var/named named]# ll named named 195 Jul 3 named named 2769 Jul 3 named named 433 Jul 3 root root 763 Oct 2 named]# chown named * named]# chgrp named * named]# ll named named named named named named named named named]# 195 2769 433 763 Jul Jul Jul Oct 3 3 3 2 2001 2001 2001 16:23 localhost. So in this case the local network (192. Once the ACL was defined. The example home network doesn’t need to have the ACL statement at all as the built in ACLs localnets and localhost are sufficient.local my-site. “Linux Networking. require a correctly configured /etc/hosts file even though DNS is correctly configured.zone 2001 2001 2001 16:23 localhost.ca named. first make sure your file permissions and ownership are okay in the /var/named directory: [root@bigboy [root@bigboy total 6 -rw-r--r-.17.zone named. Some programs.1 -rw-r--r-.0 /24).1.1 [root@bigboy [root@bigboy [root@bigboy total 6 -rw-r--r-.168.17. Remember. I then inserted a reference to the trusted-subnet in the match-clients statement in the internal view.zone The configuration files above will not be loaded until you issue the proper command to restart the named process that controls DNS. I included an ACL for network 192. and localhost get DNS data from the zone files in the internal view. trusted-subnet Loading Your New Configuration Files To load your new configuration files.20Harrison_ch18. such as sendmail.0 /24 called to help clarify the use of ACLs in more complex environments.168. this is purely an example. the other trusted network (192.

create a second NAT entry on your firewall and then create the second name server record entry with the new IP address. This screen will prompt you for the server name only.26. Sometimes. which is a private IP address. such as VeriSign and RegisterFree.100. It normally takes about three to four days for your updated DNS information to be propagated to all 13 of the world’s root name servers. namely 97. You’ll have to use NAT for Internet users to be able to gain access to the server via the chosen public IP address. Domain registrars. You’ll most likely want to test your new DNS server. .158. The actual IP address of the server is 192.158.1.my-web-site.253. Assign ns. the registrar requires at least two registered name servers per domain. If you only have one. plus a few well-known ones.26 to map to ns. You’ll therefore have to wait about this amount of time before starting to notice people hitting your new Web site.253.168. and different name.) 2. or you could give your Web server a second IP address using an IP alias. usually provide a Web interface to help you manage your domain. you may want to consider taking a look at Chapter 14.20Harrison_ch18. which should be up to date.26 in this case).qxd 2/25/05 10:09 AM Page 293 Configuring a Regular Name Server 293 Configure Your Firewall The sample network assumes that the BIND name server and Apache Web server software run on the same machine protected by a router/firewall.org.org or www. then you could either create a second name server record entry with the same IP address. but different name. Create a new name server record entry for the IP address 97. Once you’ve logged in with the registrar’s username and password.my-web-site. If your firewall is a Linux box.158.org to handle your domain.253. Fix Your Domain Registration Remember to edit your domain registration for my-web-site.” which describes how to do the network address translation and allow DNS traffic through to your name server. You can use the chapter’s troubleshooting section to test specific DNS servers for the information they have on your site.org or whatever your name server is called. or whatever it is. which should have delayed values. (This screen prompts you for both the server’s IP address and name. “Linux Firewalls Using iptables. you’ll have to take two steps: 1.my-web-site. so that at least one of the name servers is your new name server (97.

0.key [root@bigboy tmp]# To correctly relocate the files: 1.conf /var/named/chroot/etc/ 2.20Harrison_ch18.conf [root@bigboy tmp]# cat /var/named/chroot/etc/rndc. [root@bigboy tmp]# service named restart Stopping named: rndc: connect failed: connection refused . Nov 9 17:35:41 bigboy named[1157]: named -t /var/named/chroot Nov 9 17:35:41 bigboy named[1157]: Nov 9 17:35:41 bigboy named[1157]: from ‘/etc/named. Here is a sample of the /var/log/messages file that shows that named doesn’t load them:.71#53 Nov 9 17:35:41 bigboy named[1157]: on 127.2.0. Here’s a quick list of symptoms that indicate your files may not be located correctly. 127.3 -u using 1 CPU loading configuration listening on IPv4 interlistening on IPv4 intercommand channel listening command channel listening running ☞ Restarting named gives rndc “connect failed” messages: [root@bigboy tmp]# service named restart Stopping named: rndc: connect failed: connection refused [ OK ] Starting named: [ OK ] [root@bigboy tmp]# ☞ The chroot files are empty: [root@bigboy tmp]# cat /var/named/chroot/etc/named.1#953 Nov 9 17:35:41 bigboy named[1157]: on ::1#953 Nov 9 17:35:41 bigboy named[1157]: starting BIND 9.qxd 2/25/05 10:09 AM Page 294 294 Configuring DNS Chapter 18 TROUBLESHOOTING BIND One of the most common culprits of BIND problems is incorrectly located chroot files. it should shutdown correctly without error the second time.* /var/named/chroot/etc/ [root@bigboy tmp]# cp /etc/named. 10.0. ☞ The named daemon starts without loading any zone files.41. Restart named twice.conf files to the [root@bigboy tmp]# cp -f /etc/rndc.* and /var/named/chroot/etc/ directory: /etc/named. If you have the BIND chroot package installed.0. Copy the /etc/rndc.conf’ Nov 9 17:35:41 bigboy named[1157]: face lo.1#53 Nov 9 17:35:41 bigboy named[1157]: face eth0.32. make sure the configuration files are located in the chroot directory.

Check your /etc/named.0. Lack of connectivity could be caused by a firewall with incorrect permit.inaddr.172.” 2.conf file if they fail to do so. 172.” Feb 21 09:13:13 bigboy named: named startup succeeded Feb 21 09:13:13 bigboy named[12026]: loading configuration from ‘/etc/named. “Troubleshooting with syslog.in-addr.100#53 Feb 21 09:13:13 bigboy named[12026]: listening on IPv4 interface eth0.168.0.1#953 Feb 21 09:13:14 bigboy named[12026]: zone 0.20Harrison_ch18.0.168. Failure could also be caused by the named process being stopped. It is best to test this from both inside your network and from the Internet.0. (Linux logging is covered in Chapter 5. Linux status messages are logged to the file /var/log/messages.192.arpa/IN: loaded serial 51 Feb 21 09:13:14 bigboy named[12026]: zone 1. “Simple Network Troubleshooting. you can continue with the following troubleshooting steps: 1.1. 192. Check your /var/log/messages file for the loading of the zone files: Nov 9 loaded Nov 9 serial Nov 9 Nov 9 17:36:34 bigboy named[1180]: serial 1997022700 17:36:34 bigboy named[1180]: 42 17:36:34 bigboy named[1180]: 09:36:35 bigboy named: named zone 0.qxd 2/25/05 10:09 AM Page 295 Troubleshooting BIND 295 [ OK ] Starting named: [ [root@bigboy tmp]# Stopping named: [ Starting named: [ [root@bigboy tmp]# OK ] service named restart OK ] OK ] 3.arpa/IN: loaded serial 1997022700 Feb 21 09:13:14 bigboy named[12026]: zone 1.16.100#53 Feb 21 09:13:14 bigboy named[12026]: command channel listening on 127. or port forwarding rules to your DNS server.0.127.arpa/IN: zone localhost/IN: loaded running startup succeeded General Troubleshooting Steps Once your files are in the right place.in-addr. Use it to make sure all your zone files are loaded when you start BIND/named.1#53 Feb 21 09:13:13 bigboy named[12026]: listening on IPv4 interface wlan0. Troubleshooting with TELNET is covered in Chapter 4.16.0.arpa/IN: loaded serial 51 .127.inaddr. 127.1.conf’ Feb 21 09:13:13 bigboy named[12026]: no IPv6 interfaces found Feb 21 09:13:13 bigboy named[12026]: listening on IPv4 interface lo. Determine whether your DNS server is accessible on DNS UDP/TCP port 53. NAT.

com Using domain server: Name: ns1.115.my-web-site.my-web-site. Here is an example of querying DNS server ns1.org for the IP address of www.linuxhomenetworking. If this fails.org Address: 192.) .) [root@bigboy tmp]# host www.100#53 Aliases: www.20Harrison_ch18.org/IN: zone localhost/IN: loaded zone my-web-site.34 [root@bigboy tmp]# Here is an example of querying your default DNS server for the IP address of www.168.71.conf file or the Windows TCP/IP properties for your NIC.my-site. ☞ Use the dig command to determine whether the name server for your domain is configured correctly.1.linuxhomenetworking.71.115.linuxhomenetworking.linuxhomenetworking. As you can see.com.linuxhomenetworking. [root@bigboy tmp]# host www.linuxhomenetworking.com has address 65.34 [root@bigboy tmp]# ☞ You can also use the dig command to determine whether known DNS servers on the Internet have received a valid update for your zone. (Remember if you decide to change the DNS servers for your domain that it could take up to four days for it to propagate across the Internet.com ns1. the name of the specific DNS server to query has been left off the end. Failure in this case could be due not only to an error on your BIND configuration or domain registration but also to an error in your DNS client’s DNS server entry in your Linux /etc/resolv. Use the host (nslookup in Windows) command for both forward and reverse lookups to make sure the zone files were configured correctly. (You can also replace the name server’s name with its IP address.com has address 65.org/IN: running 3. try: ☞ Double check for your updated serial numbers in the modified files and also inspect the individual records within the files for mistakes. ☞ Ensure there isn’t a firewall that could be blocking DNS traffic on TCP and/or UDP port 53 between your server and the DNS server.com www.com.qxd 2/25/05 10:09 AM Page 296 296 Configuring DNS Chapter 18 Feb 21 loaded Feb 21 serial Feb 21 loaded Feb 21 09:13:14 bigboy named[12026]: serial 2004021401 09:13:14 bigboy named[12026]: 42 09:13:14 bigboy named[12026]: serial 200301114 09:13:14 bigboy named[12026]: zone my-web-site.

[root@bigboy tmp]# 3600 3600 IN IN A A 65.com domain at VeriSign. This command uses the local DNS server for the query.yahoo.net.net..net. It is sometimes good to query both your name server.70.my-isp-provider..net.my-isp-provider.yahoo. .my-isp-provider.. . linuxhomenetworking. .69 Sometimes your SOA dig will fail.net.com to make sure your DNS records have propagated properly..com linuxhomenetworking.69 Here is a successful dig using DNS server ns1... [root@bigboy tmp]# 3600 3600 IN IN A A 65. . AUTHORITY SECTION: linuxhomenetworking.. ADDITIONAL SECTION: ns1.com SOA . If you specify a name server.net. it returns the SOA record for the zone. [root@bigboy tmp]# dig ns1.com . ns1.. [root@bigboy tmp]# dig linuxhomenetworking..20Harrison_ch18...70. ns2. 3600 IN NS ns1.qxd 2/25/05 10:09 AM Page 297 Troubleshooting BIND 297 The format for the command is: dig <domain-name> <name-server> soa The name server is optional. The dig command only works with fully qualified domain names. .my-isp-provider.yahoo. 3600 IN NS ns2.com.70..com. .my-isp-provider. ns2.conf file.. because it doesn’t refer to the /etc/resolv. .my-ispprovider.68 65.115.com for the query.115.68 65.. It returns the SOA record information and the addresses of the domain’s DNS servers in the authority section. In this case the authority section doesn’t know of the domain and points to the name server for the entire .com SOA . As before.115.com.115..com. as well as a well-known name server such as ns1.my-isp-provider. ADDITIONAL SECTION: ns1.. AUTHORITY SECTION: linuxhomenetworking. linuxhomenetworking..net.. This command uses the DNS server for the query.yahoo. 3600 IN NS ns1.my-ispprovider.. then dig queries that name server instead of the Linux server’s default name server. 3600 IN NS ns2.net. .. .70. .

therefore the test server will begin to think that www. 2.org in the DNS zone file to a very low value. 0 IN SOA a. You may also want to add an entry for mail. nstld.my-web-site. ☞ Incorrect domain registration.. If anything goes wrong. . Edit the /etc/hosts file to make www. [root@bigboy tmp]# Possible causes of failure include: ☞ Typographical errors. you can then revert to the old configuration.qxd 2/25/05 10:09 AM Page 298 298 Configuring DNS Chapter 18 [root@bigboy tmp]# dig .. it will take only one minute to see the results of the final DNS configuration switch to your new server.. In this case the misspelling “linuxhomeqnetworking.gtld-servers. There is no magic bullet that will allow you to tell all the caching DNS servers in the world to flush their caches of your zone file entries. QUESTION SECTION: . Set up your test server in-house.org refer to its own IP address. At the very least your plan should include these steps: 1. ☞ A firewall could be blocking DNS traffic on TCP and/or UDP port 53 between your server and the DNS server.com..org is really hosted on itself.myweb-site. AUTHORITY SECTION: com. ns1. but there is a lag in the propagation of the domain information across the Internet.net. it will take at least three to five days for all remote DNS servers to recognize the change.verisign-grs. knowing it will rapidly recover within minutes rather than days. IN SOA . not that of the www.my-web-site.org site that is currently in production. say one minute.org if the new Web server is going to also be your new mail server. ...my-website.yahoo. 1077341254 1800 900 604800 900 .. Your best alternative is to request your existing service provider to set the TTL on my-web-site. .. ☞ Correct domain registration. Once the propagation is complete.com” was entered on the command line. As the TTL is usually set to a number of days.com.com linuxhomeqnetworking.linuxhomeqnetworking.20Harrison_ch18.. This file is usually given a higher priority than DNS.. Delays of up to four days are not uncommon.com SOA . MIGRATING YOUR WEB SITE IN-HOUSE It is important to have a detailed migration plan if you currently use an external company to host your Web site and wish to move the site to a server at home or in your office.

org map to the IP address of the new server. Test your server-based applications from the server itself. 4. You may also want to add an entry for mail.20Harrison_ch18. this could be left in the hands of your service provider. Test the server from a remote client. Edit your my-web-site. Once testing is completed. Your client will usually refer to these files first before checking DNS.org DNS entries with VeriSign. if you have concerns that your service provider won’t cooperate.my-web-site. you’ll be able to see results of the migration within minutes. you’ll need to make it assign the IP address of the Linux box as the DNS server it tells the DHCP clients to use. Once complete. In the case of Windows.my-website. You can then decide whether the change will be permanent once you have failed over back and forth a few times.my-web-site. “Configuring the DHCP Server. As the TTLs were set to one minute previously.org to point to your new Web server. you can set the TTL back to the original value to help reduce the volume of DNS query traffic hitting your DNS server. This should include mail.my-web-site. then you could explain to the provider that you want to test its failover capabilities to a duplicate server that you host in-house. Just edit your /etc/hosts file on your Web browsing Linux PC to make www. 7. DHCP CONSIDERATIONS FOR DNS If you have a DHCP server on your network. you don’t have to host DNS or mail in-house. . Web. You can then migrate these services inhouse as your confidence in hosting becomes greater. Finally.org even though DNS hasn’t been updated. then you may need to refer to Chapter 8. the file would be C:\WINDOWS\system32\drivers\etc.” SIMPLE DNS SECURITY DNS can reveal a lot about the nature of your domain. 5. You should take some precautions to conceal some of the information for the sake of security. You can test the server running as www. RegisterFree or whoever you bought your domain from to point to your new DNS servers. Remember. If your Linux box is the DHCP server. coordinate with your Web hosting provider to update your domain registration’s DNS records for www. You may also want to take over your own DNS. and so on.org if the new Web server is going to also be your new mail server. Fix your /etc/hosts files by deleting the test entries you had before. hence you can use them to predefine some DNS lookups at the local client level only. 6.qxd 2/25/05 10:09 AM Page 299 Simple DNS Security 299 3. 8.

org.168.168.100 gateway. XFR size: 16 records [root@smallfry tmp]# This may not seem like an important security threat at first glance. ntp. 3600 IN A 192.. 3600 IN A 127. In this example.conf file.. WHEN: Sun Nov 14 20:21:07 2004 . 3600 IN CNAME www.1 localhost.my-website.org. 3600 IN CNAME www.168. 3600 IN CNAME www. Anyone can use this command to determine all your server’s IP addresses and from the names determine what type of server it is and then launch an appropriate cyber attack.98 bigboy.org.org. 3600 IN MX 10 mail. You can do this by applying the allow-transfer directive to the global options section of your named.org AXFR .168.my-web-site. but it is.org. 3600 IN A 192.org.102 www.0.org. 3600 IN A 192.my-website.org. the AFXR zone transfer parameter is used to get the contents of the my-web-site.my-web-site.1 mail.168.100) .168.org. .97 192-168-1-98..3 <<>> my-web-site.100 my-web-site.1.1.1. options { allow-transfer {none.org. 3600 IN NS ns1.1.org.my-website.org.org. Query time: 16 msec . When given the right parameters it can download the entire contents of your domain’s zone file.org. smallfry.my-web-site. 3600 IN A 192. hostmaster.my-web-site.168.1. 2004110701 3600 3600 3600 3600 my-web-site. without master and slave servers. 192-168-1-96.1.org zone file: [root@smallfry tmp]# dig my-web-site. zone transfers should be disabled.org.org. 3600 IN SOA www.my-website.0.org. 3600 IN A 192.100#53(192.org. 2004110701 3600 3600 3600 3600 .my-web-site.my-website.my-web-site. ns1.168.my-web-site.org. 3600 IN A 192.my-web-site.qxd 2/25/05 10:09 AM Page 300 300 Configuring DNS Chapter 18 Zone Transfer Protection The host command does one DNS query at a time.1. }.org.1. SERVER: 192.org.my-web-site.org AXFR . In a simple home network. <<>> DiG 9.my-website.168. but the dig command is much more powerful.20Harrison_ch18..2.my-web-site.my-web-site.96 192-168-1-97.org. hostmaster.my-web-site.1. my-web-site.org. global options: printcmd my-web-site.}. 3600 IN SOA www.my-website.my-web-site..org. 3600 IN A 192.

qxd 2/25/05 10:09 AM Page 301 Conclusion 301 Once applied. or that your firewall doesn’t have the letters “FW” in its name either. There are situations in which a server’s IP address will change unpredictably and frequently.org AXFR . .org AXFR ..2. although they can be a little complicated. but to the Internet these names provide rapid identification of the types of malicious exploits a hacker could use to break in. Transfer failed. This may be good for ease of reference within the company.org domain will probably have a www and a mail subdomain. . “Dynamic DNS. You may want to adjust your DNS views so that to external users. your MySQL database server doesn’t have the letters “DB” or “SQL” in the name. [root@smallfry tmp]# dig my-web-site. <<>> DiG 9.. because the IP address of a server is normally fixed or static. Dynamic DNS was created as a solution to this and is explained in Chapter 19. global options: printcmd . This is not always the case. Web site security refers to anything that helps to guarantee the availability of the site. [root@smallfry tmp]# Naming Convention Security Your my-web-site.. making DNS management extremely difficult.3 <<>> my-web-site. Fortunately.. your zone transfer test should fail..20Harrison_ch18. DNS modifications are usually infrequent. and they should remain obvious to all.” . CONCLUSION DNS management is a critical part of the maintenance of any Web site. this is just one of many methods you can use.

qxd 2/25/05 10:09 AM Page 302 .20Harrison_ch18.

the DSL or cable modem IP address is provided by DHCP and therefore changes from time to time. This chapter will explain the details of dynamic DNS configuration.org domain.qxd 2/25/05 10:09 AM Page 303 C H A P T E R 19 Dynamic DNS IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ Dynamic DNS Preparation Dynamic DNS and NAT Router/Firewalls DDNS Client Software: SOHO Router/Firewalls DDNS Client Software: Linux DDclient Testing Your Dynamic DNS Conclusion In many home networking environments.21Harrison_ch19.org domain. You will have to use the services of a third-party DNS provider to provide DNS information for your my-web-site.” which assume the IP address of a server won’t change continuously. ☞ Dynamic DNS (DDNS): Used when you get a changing dynamic IP addresses via DHCP from your ISP. For this reason there are two broad types of DNS: ☞ Static DNS: This is used when your ISP provides you with unchanging fixed or static IP addresses. You can consider static DNS as the “traditional” or “regular” form of DNS. “Configuring DNS. 303 . Your DNS server acts as the authoritative source of information for your my-web-site. This can cause problems with the DNS zone files explained in Chapter 18.

it has disadvantages.) 5. Make sure you can ping your default gateway. Use the ipconfig command to determine the new IP address of your Web server.21Harrison_ch19. Ask a friend somewhere else on the Internet to try to TELNET to this IP address on port 80. “The Apache Web Server. (This command is explained in Chapter 3 also. then ask the friend to point a Web browser to the IP address and see whether a valid Web page appears. 4. Setup You need to do some basic setup before testing can begin.” 2.qxd 2/25/05 10:09 AM Page 304 304 Dynamic DNS Chapter 19 DYNAMIC DNS PREPARATION Unlike DSL. and configure the Ethernet NIC for DHCP as described in Chapter 3.” 3. Web queries first hit their servers on the regular TCP ports and then these servers automatically redirect the Web clients to use the IP address of your server on a different TCP port.” If you can get through to the IP address on port 80. Testing ISP Connectivity for Your Web Site The very first thing you need to do is to determine whether your ISP allows inbound connections on your DSL or cable modem line. Connect your Linux server directly to your cable or DSL modem. . and many search engines do not index redirected pages. most cable modem providers may not allow you to host sites at home by blocking inbound HTTP (TCP port 80) and SMTP mail (TCP port 25) while allowing most other TCP traffic through. From the Linux Web server itself. try to TELNET to this IP address on port 80 as explained in Chapter 4. but in some cases they’ll say no when in fact the answer is yes. “Simple Network Troubleshooting. Configure and start Apache on your Linux Web server as described in Chapter 20. Here is how you can test it out for yourself. then you most likely have Apache configured correctly Testing from the Internet You may be able to see Web pages from the Web server itself. 1. The cost of the service can make hosting with a $10 /month virtual hosting service look very attractive. If your Linux server will eventually be placed behind a firewall. Under the system. The easiest way to do this is to phone them and ask. Although this works well. “Linux Networking. If port 80 works. then adjust your network topology accordingly and test port 80 port forwarding to your Web server. Many DDNS providers are aware of this and provide a redirect service to bypass the problem.

then adjust your network topology accordingly. In this example. then you can additionally sign up for a redirect service with your DDNS provider as explained earlier.org/ | grep -i address <html><head><title>Current IP Check</title></head><body>Current IP Address: 24.21Harrison_ch19. you are interested in only the line that tells you what the server thinks your IP address is.110: [root@bigboy tmp]# curl -s http://checkip. If it works. the IP address is 24. You can determine the external IP address of your router or firewall by logging onto your Linux Web server and issuing the curl command to query the DynDNS.4.qxd 2/25/05 10:09 AM Page 305 Dynamic DNS Preparation 305 If the TELNET to port 80 fails. remember to adjust the rules to match this new port or stop iptables temporarily while doing this testing. In this case.conf listen directive.dyndns.97.org IP information server. try the Web browser test too. Listen 80 Test again with TELNET on this new port. Let Apache run on port 80. preferably above 1024.110</body></html> [root@bigboy tmp]# Registering DDNS Once you decide to go ahead with DDNS. Test Port Forwarding If your Linux server will eventually be placed behind a firewall. you need to choose between the broad categories of Dynamic DNS service: . If you can get a connection with correctly displayed pages on a nonstandard port. If the test port is 1234. change “80” to some other value. One of the features of the curl troubleshooting utility is that it allows you to download a Web page’s HTML code to your screen without interpreting it. Change it to your new value and restart httpd: # httpd. use the grep command to extract just that part. Note If you are running iptables. then use a URL of http://server-ipaddress:1234. and test port forwarding from the nonstandard port to port 80 on your Web server from the Internet.conf file uses the listen directive to do this.4.97. The Apache httpd. your ISP probably doesn’t allow HTTP access to its networks. Configure your Web server to run on a different TCP port.

now comes in a RPM format.net. 3. When you own your own domain this worry largely goes away as you are fully in control of the DNS registration and renewal process. and register your Web sites (sometimes called hosts) as part of your domain (as in www.org) with them. Your DDNS registration process will provide you with a username and password.my-isp. The very popular one DDclient. for example. Another disadvantage is that you must rely on your DDNS provider staying in business or else you may lose your domain.org and mail.my-website. You can perform all the necessary steps on your DDNS service provider’s Web site.org to the DNS name servers of the DDNS provider. then your site will become my-site. Install a DDNS Client on Your Server All DDNS service providers require that you use a DDNS client on your Web server that will periodically update the IP address information in your provider’s DDNS record.net.my-web-site. If you choose to create your own domain and use a paid DDNS service. Update your domain information with your main DNS registrar (Verisign and RegisterFree) to tell them to direct queries to *.my-website. You should also be prepared for slower response times for your homebased site than if you were using a static IP and a regular DNS service. which you’ll need to use when configuring your DDNS client.org) with such companies as Verisign and RegisterFree. if the DDNS provider’s domain is isp. 4. follow these steps: 1.21Harrison_ch19. ☞ Paid Customized DNS: You can register the domain name of your choice and still host your Web site on a DHCP line. Register your domains (such as my-web-site. 2. Install a DDNS client on your Web servers that continuously runs. Create an account with the DDNS provider. only updating the DDNS provider’s DNS servers with the most current DHCP IP address of the site whenever it detects a change. . Remember that this type of service may be undesirable for a company that wants to establish its own corporate identity. For example.qxd 2/25/05 10:09 AM Page 306 306 Dynamic DNS Chapter 19 ☞ Free Dynamic DNS: Your Web site name will be a subdomain of the DDNS provider’s domain.

“Installing RPM Software. then the examples in Chapter 14.21Harrison_ch19. Some DDNS providers use more intelligent clients. . then the NIC reports in its data stream to the DDNS provider a private IP address that no one can reach directly via the Internet. that can be configured to let the DDNS provider record the public IP address from which the data stream is originating. If the Linux box is being protected behind a NAT router/firewall.noarch.” will be helpful. the software can report only the true IP address of the Linux box’s NIC interface. “Linux Firewalls Using iptables. The reported value is therefore invalid. “Introduction to Networking. The RPM name usually starts with ddclient followed by a version number.qxd 2/25/05 10:09 AM Page 307 DDNS Client Software: Linux DDclient 307 DYNAMIC DNS AND NAT ROUTER/FIREWALLS As discussed in Chapter 2. Like most Red Hat and Fedora Linux software products.3-1. With this support. There is usually a Dynamic DNS Web menu that prompts for the name of the service provider and your DDNS username and password. DHCP-obtained IP address to the many private IP addresses within your network. which can overcome the NAT limitations of DDNS by actually logging into your SOHO firewall to determine the latest IP address information. public. Once this is done. (If you need a refresher on RPMs.net. DDNS CLIENT SOFTWARE: LINUX DDCLIENT One of the most commonly used clients is DDclient. If your firewall is Linux based. DDclient is available in the RPM format. such as DDclient. consult Chapter 6. NAT can fool the operation of some DDNS client software.6. DDNS CLIENT SOFTWARE: SOHO ROUTER/FIREWALLS Most new SOHO router/firewalls have built in dynamic DNS clients for one or more of the major DDNS service providers. as in ddclient-3.”) You can usually download the software from your DDNS provider.” to conserve the limited number of IP addresses available for Internet purposes. most home router/firewalls use network address translation (NAT) to map a single. or you can find it at rpmfind. you’ll have to also configure your router/firewall to do port forwarding to make all HTTP traffic destined for the IP address of the router/firewall to be exclusively address translated using NAT and forwarded to a single server on your home network. Many Web-based small office/home office (SOHO) firewalls have easy interfaces to configure port forwarding. please refer to your manufacturer’s manual on how to do this. In these cases. there is no need to install software on your Web server.rpm.

100 use=web.dnsalias.conf file.168.com \ \ You can add one of the following use lines to the General Parameter Section near the top of the file to define the method that will be used to determine the correct IP address: . In this example. protocol=dyndns2 mysite-example.qxd 2/25/05 10:09 AM Page 308 308 Dynamic DNS Chapter 19 The /etc/ddclient. The most important general parameters to configure are: ☞ Username: Your DDNS account’s login name ☞ Password: Your DDNS account’s password ☞ Use: The method used to determine the IP address to advertise to the DDNS server Some important DDNS provider parameters are: ☞ Server: The name of the DDNS provider’s main DNS server ☞ Protocol: The methodology the DDNS client should use to communicate with the DDNS server ☞ Your domain: The domain to which your Web server will belong You can use the ddclient command to determine the best use parameter to use in the ddclient.org. if=wlan0 address is 192.158.dnsalias.253.com. if=lo address is 127.1 use=if. This example specifies a username of my-account-login-name and a password of my-account-password using the dyndns DDNS service provider’s settings to track the Web site named mysite-example.0.conf File The ddclient.21Harrison_ch19. web=dyndns address is 97.0.26 [root@bigboy tmp]# Take a look at a sample configuration.dyndns.conf file is usually installed completely commented out. # General Parameter Section login=my-account-login-name password=my-account-password # DDNS Provider Parameters Section server=members.1. but provides many configuration examples for the most popular DDNS providers. only the use=web option gives a valid Internet IP address and should be considered as a first option: [root@bigboy tmp]# ddclient -daemon=0 –query use=if.

qxd 2/25/05 10:09 AM Page 309 DDNS Client Software: Linux DDclient 309 ☞ Query a well-known Internet server: The Web method queries two well known servers run by DynDNS. This method is the simplest as it requires no further information and handles NAT correctly. use=if. you should get a response of plain old process ID numbers: [root@bigboy tmp]# pgrep ddclient . How to Get DDclient Started You can configure DDclient to start at boot time using the chkconfig command: [root@bigboy tmp]# chkconfig ddclient on You can start. and not via NAT.21Harrison_ch19.conf file for the changes to take effect on the running process. you can create your own parameters as outlined in the DDclient read-me file. This option is good for NAT environments where the use=web option isn’t considered a good alternative. You can test whether the DDclient process is running with the pgrep command. as in: [root@bigboy tmp]# service ddclient start [root@bigboy tmp]# service ddclient stop [root@bigboy tmp]# service ddclient restart Remember to restart the DDclient process every time you make a change to the ddclient. and restart DDclient after boot time using the DDclient initialization script.org and DNSpark to determine the public Internet IP address of the Web server running the DDclient software. This is probably most valuable for servers connected directly to the Internet. use=web ☞ Use the IP address of a specific server NIC: You can also query the IP address of the DDclient Web server’s NIC interface of your choice. If your model isn’t listed.conf file has a list of use statements for various vendor’s firewalls. stop. After editing your configuration file. you’ll have to start ddclient as shown in the next section. if=eth0 ☞ Login to your SOHO firewall for information: The ddclient.

my-web-site...conf files are good sources of information for doing custom configurations and troubleshooting.0.6. you can use the files: locate command to find all the DDclient [root@bigboy tmp]# locate ddclient | grep READ /usr/doc/ddclient-3.0.qxd 2/25/05 10:09 AM Page 310 310 Dynamic DNS Chapter 19 Finding DDclient Help The DDclient read-me and ddclient. one of which is the read-me file: [root@bigboy tmp]# rpm -ql ddclient .. [root@bigboy tmp]# Alternately. you will get an error message like: [root@bigboy tmp]# host www.my-web-site.3/README .cisco [root@bigboy tmp]# locate -u If the command doesn’t work..0.6.1#53 ** server can’t find www. Use the rpm command to get a list of installed DDclient files.3/README. /usr/doc/ddclient-3. You can find the read-me file by using DDclient RPMs or the locate command.1 Address: 127.6.21Harrison_ch19.org command from your Linux command prompt to determine whether you are getting a valid response.0. ☞ Using the host www. try updating your locate database with the command followed by the locate ddclient command once more: [root@bigboy tmp]# locate -u TESTING YOUR DYNAMIC DNS You can test your dynamic DNS by: ☞ Looking at the status page of your DNS provider and making sure the IP address that matches your www site is the same as your router/firewall’s public IP address.my-web-site. If you failed to add your host record.org: NXDOMAIN .org Server: 127.3/README /usr/doc/ddclient-3.

my-web-site. . The example below queries the miniDNS name server ns1. you may need to consider a regular data center far away from spilt coffee and the washing machine that always trips the circuit breakers.158.my-web-site.21Harrison_ch19. maintained.org ns1. but it is frequently unreliable as residential class broadband data circuits are not monitored. but as your Web site becomes busier and more financially important to you. It is a good starting place to help you become familiar with Web hosting.qxd 2/25/05 10:09 AM Page 311 Conclusion 311 This error could be because your domain hasn’t propagated fully throughout the Internet.minidns.net www.253.org has address 97.net: [root@bigboy tmp]# host www.minidns. You can test to make sure everything is okay by forcing NS lookup to query the name servers directly.26 [root@bigboy tmp]# CONCLUSION Always remember that dynamic DNS works. or managed as closely as business class lines.

qxd 2/25/05 10:09 AM Page 312 .21Harrison_ch19.

This chapter explains how to configure Apache in a number of commonly encountered scenarios for small Web sites.2. as in httpd-2. see Chapter 6.rpm.48-1. Once you have DNS correctly set up and your server has access to the Internet. DOWNLOADING AND INSTALLING THE APACHE PACKAGE Most Red Hat and Fedora Linux software products are available in the RPM format.22Harrison_ch20. “Installing RPM Software. remember that the Apache RPM’s filename usually starts with the word httpd followed by a version number. (For more on RPMs. It is best to use the latest version of Apache.qxd 2/25/05 10:11 AM Page 313 C H A P T E R 20 The Apache Web Server IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Downloading and Installing the Apache Package Starting Apache Configuring DNS for Apache DHCP and Apache General Configuration Steps Configuration: Multiple Sites and IP Addresses Using Data Compression on Web Pages Apache Running on a Server Behind a NAT Firewall Protecting Web Page Directories with Passwords The /etc/httpd/conf.0. you’ll need to configure Apache to accept surfers wanting to access your Web site.” 313 .d Directory Troubleshooting Apache Conclusion Apache is probably the most popular Linux-based Web server application in use. When searching for the file.

(Chapter 19 explains port forwarding. as well. or static IP address that is outside of the range of the DHCP server to prevent you from having this problem.1.” DHCP AND APACHE As you remember. This could cause your firewall port forwarding. the Web server lives on the 192.0 network uses a fixed.qxd 2/25/05 10:11 AM Page 314 314 The Apache Web Server Chapter 20 STARTING APACHE Use the chkconfig command to configure Apache to start at boot: [root@bigboy tmp]# chkconfig httpd on Use the httpd init script in the restart Apache after booting: /etc/init. then you probably also need to set up port forwarding for your Web traffic to reach the Web server correctly.d directory to start. to break. if your Internet connection uses DHCP to get its IP address.” or Chapter 19. . refer to Chapter 18. CONFIGURING DNS FOR APACHE Remember that you will never receive the correct traffic unless you configure DNS for your domain to make your new Linux box Web server the target of the DNS domain’s www entry. then you need to use dynamic DNS to get the correct Internet DNS entry for your Web server. I recommend that your Web server on the 192.0 home network protected by a firewall. and [root@bigboy tmp]# service httpd start [root@bigboy tmp]# service httpd stop [root@bigboy tmp]# service httpd restart You can test whether the Apache process is running with: [root@bigboy tmp]# pgrep httpd You should get a response of plain old process ID numbers. Remember that the IP address of your Web server can change if it gets its IP address using DHCP. “Dynamic DNS. The firewall uses NAT and port forwarding to pass Internet traffic on to the Web server.1. In this case. In the book’s sample topology. “Configuring DNS. stop.168. To do this.22Harrison_ch20. If your Web server and firewall are different machines.168. not dynamic DNS.) DHCP on your protected home network is different.

index. ☞ DocumentRoot: Defines the directory in which the Web pages for the site can be found. Apache searches the DocumentRoot directory for an index.htm with index.html pointing to the file index. as I’ll explain soon.org with a DocumentRoot directory of /home/www/site1/.html.conf. The easiest solution is to create a symbolic link (known as a shortcut to Windows users) called index.htm page.html in the /home/www/site1 directory: [root@bigboy [root@bigboy [root@bigboy -rw-rw-r-lrwxrwxrwx index. This is needed in Web site managed by the named virtual hosting only. such as Microsoft FrontPage.html -> site1]# The l at the very beginning of the the -> indicates the link target. You’ll almost never have to worry about index. Apache displays the contents of the file /home/www/site1/index.my-web-site.html being updated automatically.htm [root@bigboy tmp]# cd /home/www/site1 site1]# ln -s index. or home. or directives. The most commonly used statements. So for example.htm as FrontPage does. if you have a servername of www. page named index. or container.htm extension. This isn’t usually a problem if all your HTML files have hyperlinks pointing to files ending in .* 1 root root 48590 Jun 18 23:43 index. create files with an . you must restart Apache before changes to this configuration file take effect.22Harrison_ch20.html entry signifies a link and . Where to Put Your Web Pages All the statements that define the features of each Web site are grouped together inside their own <VirtualHost> section. Some editors.html when you enter http://www.org in your browser.my-web-site. inside a <VirtualHost> container are: ☞ servername: Defines the name of the <VirtualHost> container.htm index. The problem occurs with Apache not recognizing the topmost index.qxd 2/25/05 10:11 AM Page 315 General Configuration Steps 315 GENERAL CONFIGURATION STEPS The configuration file used by Apache is /etc/httpd/conf/httpd.html. By default.conf file.html site1]# ll index. As for most Linux applications. not .htm. This then enables you to edit or copy the file index.htm 1 root root 9 Jun 21 18:05 index. in the httpd.html and Apache again! This example creates a symbolic link to index.

Scripts. It is a good idea to have the files owned by a nonprivileged user so that Web developers can update the files using FTP or SCP without requiring the root password.* -rw-rw-r-1 www users 48590 Jun 25 23:43 index. Recursively change the file ownership permissions of the /home/www directory and all its subdirectories. This will make all the transferred files automatically have the correct ownership. and Configurations.conf. Change the permissions on the /home/www directory to 755. You have to make sure all the files and subdirectories in your DocumentRoot have the correct permissions. You may also have to use the Directory directive to make Apache . 2.html -> index.qxd 2/25/05 10:11 AM Page 316 316 The Apache Web Server Chapter 20 The Default File Location By default.htm [root@bigboy tmp]# Tip Be sure to FTP or SCP new files to your Web server as this new user. If you browse your Web site after configuring Apache and get a “403 Forbidden” permissions-related error on your screen.” has a short script that you can use to recursively set the file permissions in a directory to match those expected by Apache. Apache expects to find all its Web page files in the /var/www/html/ directory with a generic DocumentRoot statement at the beginning of httpd. then your files or directories under your DocumentRoot most likely have incorrect permissions. The examples in this chapter use the /home/www directory to illustrate how you can place them in other locations successfully. File Permissions and Apache Apache will display Web page files as long as they are world readable. including the Apache’s httpd daemon. The code you need is: [root@bigboy tmp]# useradd -g users www [root@bigboy tmp]# chown -R www:users /home/www [root@bigboy tmp]# chmod 755 /home/www Now test for the new ownership with the ll command: [root@bigboy tmp]# ll /home/www/site1/index.22Harrison_ch20. Create a user with a home directory of /home/www. “Codes. To do this: 1. 3. to read the files inside. Appendix II.htm lrwxrwxrwx 1 www users 9 Jun 25 18:05 index. which allows all users.

1 SELinux Security Context File Labels Description The type used by regular static Web pages with . Contextrelated configuration errors will give “403 Forbidden” browser messages. but also allows appending data to files by the CGI script. then they will inherit the context of that directory and you should have very few problems. Security Contexts for Web Pages Fedora Core 3 introduced the concept of security contexts as part of the Security Enhanced Linux (SELinux) definition. The context of a file depends on the SELinux label it is given. Required for CGI scripts to read files and directories. In this example. security contexts become important when Web pages need to be placed in directories that are not the Apache defaults. If you decide to place your Web pages in the default /var/www/ directory. The most important types of security label are listed in Table 20. If you have your files in the default /home/www directory.html page created in it.html [root@bigboy tmp]# ls -Z /home/www/site1/index. Files with this type may be changed by a CGI script in any way. it inherits the security context of its parent directory. then this second step becomes unnecessary.html [root@bigboy tmp]# site1 . including deletion.qxd 2/25/05 10:11 AM Page 317 General Configuration Steps 317 serve the pages once the file permissions have been correctly set. “Miscellaneous Linux Topics. When a file is created.html and . and in some cases.) A Web page may have the right permissions.22Harrison_ch20. HTTP Code httpd_sys_content_t httpd_sys_script_ro_t httpd_sys_script_ra_t httpd_sys_script_rw_t httpd_sys_script_exec_t As expected.1. The type required for the execution of CGI scripts. (See Appendix I.html -rw-r--r-.root root root:object_r:user_home_t /home/www/site1/index. Table 20. you can see that the user_home_t security label has been assigned to the directory and the index. user root creates a directory /home/www/site1 in which the pages for a new Web site will be placed. Using the ls -Z command. This label is not accessible by Apache.” for details. Same as the httpd_sys_script_ro_t type. but the Apache httpd daemon won’t be able to read it unless you assign it the correct security context or daemon access permissions. [root@bigboy tmp]# mkdir /home/www/site1 [root@bigboy tmp]# ls -Z /home/www/ drwxr-xr-x root root root:object_r:user_home_t [root@bigboy tmp]# touch /home/www/site1/index. you will get the default Fedora Apache page where your expected Web page should be.htm extensions.

[root@bigboy tmp]# tail /var/log/httpd/error_log [Fri Dec 24 17:59:24 2004] [error] [client 216. CGI scripts can be written in a variety of languages.qxd 2/25/05 10:11 AM Page 318 318 The Apache Web Server Chapter 20 Accessing the index.root root root:object_r:httpd_sys_content_t index. A Web page’s Submit button usually has a CGI script lurking somewhere beneath.txt is created: chcon [root@bigboy tmp]# touch /home/www/site1/test.root root root:object_r:httpd_sys_content_t index. Viewing the /var/log/httpd/error_log gives a “Permission Denied” message and the /var/log/ messages file shows kernel audit errors.txt [root@bigboy tmp]# Security Contexts for CGI Scripts You can use Apache to trigger the execution of programs called Common Gateway Interface (CGI) scripts. CGI scripts are placed in the /var/www/cgi-bin/ directory as defined by the .html [root@bigboy tmp]# Browsing now works without errors.root root root:object_r:httpd_sys_content_t test. Recognizing the error.250] (13)Permission denied: access to / denied [root@bigboy tmp]# tail /var/log/messages Dec 24 17:59:24 bigboy kernel: audit(1103939964.html -rw-r--r-.10. [root@bigboy tmp]# chcon -R -h -t httpd_sys_content_t /home/www/site1 [root@bigboy tmp]# ls -Z /home/www/site1/ -rw-r--r-. and can be used to do such things as generate new Web page output or update data files. By default.119.html file via a Web browser gets a “Forbidden 403” error on your screen. user root uses chcon with the -R (recursive) and -h (modify symbolic links) qualifiers to modify the label of the directory to httpd_sys_content_t with the -t qualifier. You can see this when the file /home/www/site1/test. User root won’t have to run the command again for the directory. because new files created in the directory will inherit the SELinux security label of the parent directory. even though the permissions are correct. including PERL and PHP.444:0): avc: denied { getattr } for pid=2188 exe=/usr/sbin/httpd path=/home/www/site1 dev=hda5 ino=73659 scontext=system_u:system_r:httpd_t tcontext=root:object_r:user_home_t tclass=dir [root@bigboy tmp]# SELinux security context labels can be modified using the chcon command.22Harrison_ch20.txt [root@bigboy tmp]# ls -Z /home/www/site1/ -rw-r--r-.

1 for a CGI script to be run in another directory or to access data files. In the example case.cgi” print qq( <html> <head> <meta http-equiv=”Content-Language” content=”en-us”> <meta http-equiv=”Content-Type” content=”text/html”> <title>Linux Home Networking</title> </head> <body> <p align=”center”><font size=”7”>Success!</font></p> </body> </html> ).cgi was created to display the word “Success” on the screen of your Web browser. any URL with the string /cgi-bin/ will trigger Apache to search for an equivalent executable file in this directory. SELinux contexts have to be modified according to the values in Table 20. changes the directory’s security context label to httpd_sys_script_exec_t.cgi with the correct executable file permissions. The ScriptAlias directive was set to point to of /var/www/cgi-bin/. [root@bigboy tmp]# mkdir -p /home/www/cgi-bin/test [root@bigboy tmp]# chcon -h -t httpd_sys_script_exec_t /home/www/cgi-bin/ [root@bigboy tmp]# mkdir /home/www/cgi-bin/test [root@bigboy tmp]# ls -Z /home/www/cgi-bin drwxr-xr-x root root root:object_r:httpd_sys_script_exec_t test [root@bigboy tmp]# vi /home/www/cgi-bin/test/test. and then creates the script /home/www/cgi-bin/test/test.22Harrison_ch20.cgi actually executes the script file /var/www/cgi-bin/test/test.1.cgi [root@bigboy tmp]# .qxd 2/25/05 10:11 AM Page 319 General Configuration Steps 319 ScriptAlias directive you’ll find in the more detail later.cgi. for example. the PERL script test. the URL http://192.168. #!/usr/bin/perl # CGI Script “test. ScriptAlias /cgi-bin/ “/home/www/cgi-bin/” /home/www/cgi-bin/ instead User root creates the /home/www/cgi-bin/ directory. which I’ll discuss in ScriptAlias /cgi-bin/ “/var/www/cgi-bin/” In the default case.conf file.cgi [root@bigboy tmp]# chmod o+x /home/www/cgi-bin/test/test. httpd.100/cgi-bin/test/test. So.

qxd 2/25/05 10:11 AM Page 320 320 The Apache Web Server Chapter 20 Accessing the URL http://192.119. ScriptAlias /cgi-bin/ “/web/cgi-bin/site2/” The file inherits the context of its new parent. SELinux will be disabled after your next reboot.cgi -rw-r--r-x root root root:object_r:tmp_t /web/cgi-bin/site2/test. The file is copied to a directory /web/cgi-bin/site2/ governed by the ScriptAlias in the second Web site’s <VirtualHost> container (explained later).168.cgi /web/cgibin/site2/test. you can turn them off system wide by editing your /etc/selinux/config file.cgi file.cgi file needs to be used by a second Web site housed on the same Web server.22Harrison_ch20.cgi is successful. .10.cgi denied [root@bigboy tmp]# tail /var/log/messages Dec 24 18:36:08 bigboy kernel: audit(1103942168.549:0): avc: denied { getattr } for pid=2191 exe=/usr/sbin/httpd path=/web/cgibin/site2/test.1. [root@bigboy tmp]# tail /var/log/httpd/error_log [Fri Dec 24 18:36:08 2004] [error] [client 216. You use the NameVirtualHost directive in the /etc/httpd/conf/httpd.conf file to tell Apache which IP addresses will participate in this feature.100/cgi-bin/test/test. Problems occur when the same test.cgi [root@bigboy tmp]# ls -Z /web/cgi-bin/site2/test. but the security context label isn’t copied along with it.cgi [root@bigboy tmp]# Permission denied and kernel audit errors occur once more. Named Virtual Hosting You can make your Web server host more than one site per IP address by using Apache’s named virtual hosting feature.250] (13)Permission denied: access to /cgi-bin/texcelon/test. you can fix them only by changing the security context of the test.cgi dev=hda5 ino=77491 scontext=system_u:system_r:httpd_t tcontext=root:object_r:tmp_t tclass=file [root@bigboy tmp]# Note If you find security contexts too restrictive. modifying the SELINUX parameter to disabled. [root@bigboy tmp]# cp /home/www/cgi-bin/test/test.

not site #1 or site #2) </VirtualHost> <VirtualHost 97. IP-Based Virtual Hosting The other virtual hosting option is to have one IP address per Web site.26> servername www. In this case.253.158. In the code that follows. <VirtualHost>.26> servername www.253.253. notice that within each <VirtualHost> container you specify the primary Web site domain name for that IP address with the ServerName directive. in cases where your Web server has more than one IP address.my-web-site. NameVirtualHost 97.158. The DocumentRoot directive defines the directory that contains the index page for that site. even if the ServerName directive doesn’t match.253.org Directives for site #2 </VirtualHost> Be careful with using the asterisk in other containers. try to limit the use of your <VirtualHost *> statements to the beginning of the list to cover any other IP addresses your server may have. This is why the first <VirtualHost> statement contains an asterisk: to indicate it should be used for all other Web queries. You can also have multiple NameVirtualHost directives.qxd 2/25/05 10:11 AM Page 321 General Configuration Steps 321 The <VirtualHost> containers in the file then tell Apache where it should look for the Web pages used on each Web site. If there is no match.158. (In other words. which is also known as IP-based virtual hosting.158. To get consistent results. Apache searches for a perfect match of NameVirtualHost.26 <VirtualHost *> Default Directives. then Apache uses the first <VirtualHost> in the list that matches the target IP address of the request. you will not have a .22Harrison_ch20. each with a single IP address. Named Virtual Hosting Example Consider an example in which the server is configured to provide content on 97. You can also list secondary domain names that will serve the same content as the primary ServerName using the ServerAlias directive. A <VirtualHost> with a specific IP address always gets higher priority than a <VirtualHost> statement with an * intended to cover the same IP address. You must specify the IP address for which each <VirtualHost> container applies. and ServerName when making a decision as to which content to send to the remote user’s Web browser.org Directives for site #1 </VirtualHost> <VirtualHost 97.26.another-web-site.

but gives different content for addresses 97.253. Web surfers get the site1 content if they try to access the Web server on any of its other IP addresses.158. <VirtualHost *> DocumentRoot /home/www/site1 </VirtualHost> IP Virtual Hosting Example: Wild Card and IP Addresses In this example. <VirtualHost *> DocumentRoot /home/www/site1 </VirtualHost> <VirtualHost 97. When you use wild cards. because there is only one Web site per IP address. which is used frequently in credit card and shopping cart Web pages. the ServerName directive isn’t needed in each <VirtualHost> container. If you installed Apache with support for secure HTTPS/SSL. Apache interprets it as an overlap of name-based and IP-based <VirtualHost> .27> DocumentRoot /home/www/site3 </VirtualHost> Virtual Hosting and SSL Because it makes configuration easier.22Harrison_ch20. but gives the same content. system administrators commonly replace the IP address in the <VirtualHost> and NameVirtualHost directives with the * wildcard character to indicate all IP addresses.158.253. IP Virtual Hosting Example: Single Wild Card In this example. Apache displays the content in the first <VirtualHost *> directive even if you add another right after it.27. <VirtualHost> container per IP address. and you must only have a single Also.253. Apache listens on all interfaces.158. The Apache SSL module demands at least one explicit <VirtualHost> directive for IP-based virtual hosting. unlike in named virtual hosting.qxd 2/25/05 10:11 AM Page 322 322 The Apache Web Server Chapter 20 NameVirtualHost directive for the IP address. then wild cards won’t work.26> DocumentRoot /home/www/site2 </VirtualHost> <VirtualHost 97.26 and 97.158. Apache listens on all interfaces. Apache also seems to enforce the single <VirtualHost> container per IP address requirement by ignoring any ServerName directives you may use inside it.253.

22Harrison_ch20.253.my-web-site. my-web-site.27.org to map to the IP address 97.158. Here is an example: NameVirtualHost * <VirtualHost *> Directives for other sites </VirtualHost> <VirtualHost 97. .253. The administrator wants to be able to get to test.org.my-web-site. The parameters are: ☞ The Web site’s systems administrator previously created DNS entries for www.26 on this Web server. www. proceeding with undefined results If you try to load any Web page on your Web server.conf file.158.my-web-site. If you think this is a server error.org.org on all the IP addresses. please contact the webmaster The best solution to this problem is to use wild cards more sparingly.org is also configured to point to alias IP address 97. I’ll walk you through an example scenario. Don’t use virtual hosting statements with wild cards except for the very first <VirtualHost> directive that defines the Web pages to be displayed when matches to the other <VirtualHost> directives cannot be found.28> Directives for site that also run on SSL </VirtualHost> CONFIGURATION: MULTIPLE SITES AND IP ADDRESSES To help you better understand the edits needed to configure the /etc/httpd/conf/httpd.org.qxd 2/25/05 10:11 AM Page 323 Configuration: Multiple Sites and IP Addresses 323 directives and gives error messages because it can’t make up its mind about which method to use: Starting httpd: [Sat Oct 12 21:21:49 2002] [error] VirtualHost _default_:443 -. The domain www. and default. you’ll also see the error: Bad request! Your browser (or proxy) sent a request that this server could not understand.another-web-site.253.mixing * ports and non-* ports with a NameVirtualHost address is not supported.my-web-business.158.

my-web-site.26 as in this case we have a single IP address serving different content for a variety of domains. A NameVirtualHost directive for 97.26 NameVirtualHost 97.158.27 97. Table 20.org www.org must get content from subdirectory site3. Table 20.158.26 Directory Site2 Type of Virtual Hosting Name based 97. ☞ All other domains pointing to this server that don’t have a matching ServerName directive will get Web pages from the directory defined in the very first <VirtualHost> container: directory site1.my-web-site.253.my-web-business.253.253.org.158.org www.html in this directory.158.org will get content from directory site4.my-web-business. and www.org.org test.253.158.my-web-site.253.27 # # Match a webpage directory with each website # <VirtualHost *> DocumentRoot /home/www/site1 </VirtualHost> <VirtualHost 97. ☞ Traffic to test.158.another-web-site. www. Site default.2 Domain www.253.27 97.conf file: ServerName localhost NameVirtualHost 97.my-web-site.org my-web-site. my-web-site.org falls in this category.org must get content from subdirectory site2.my-web-site.org.org </VirtualHost> .158. ☞ Named virtual hosting will be required for 97.my-website.26 Site3 Site1 Site1 Name based (Wild card) Name based Name based All other domains How do these requirements translate into code? Here is a sample snippet of a working httpd.org default.qxd 2/25/05 10:11 AM Page 324 324 The Apache Web Server Chapter 20 ☞ Traffic to www.my-web-business.253. ☞ Traffic going to www.org Web Hosting Scenario Summary IP address 97.org ServerAlias my-web-site.253. Hitting these URLs causes Apache to display the contents of index.26> DocumentRoot /home/www/site2 ServerName www.another-web-site.2 summarizes these requirements.22Harrison_ch20.my-web-site.158.253.158.26 is therefore required.

Remember to restart Apache every time you update the httpd.deny Allow from all </Limit> <LimitExcept GET POST OPTIONS> Order deny. Testing Your Web Site Before DNS Is Fixed You may not be able to wait for DNS to be configured correctly before starting your project. The last section of this configuration snippet has some additional statements to ensure read-only access to your Web pages with the exception of Web-based forms using POSTs (pages with “submit” buttons).253.org.my-web-site.158.another-web-site.27> DocumentRoot /home/www/site4 ServerName www.allow Deny from all </LimitExcept> </Directory> These statements would normally be found at the very bottom of the file where the virtual hosting statements reside.253. # <Directory “/home/www/*”> Order allow. Note You will have to configure your DNS server to point to the correct IP address used for each of the Web sites you host. .org </VirtualHost> # # Make sure the directories specified above # have restricted access to read-only.org and another-web-site.org </VirtualHost> <VirtualHost 97.158.27> DocumentRoot /home/www/site3 ServerName test.qxd 2/25/05 10:11 AM Page 325 Configuration: Multiple Sites and IP Addresses 325 <VirtualHost 97. on your DNS server. such as my-web-site. The easiest way to temporarily bypass this is to modify the hosts file on the Web developer’s client PC or workstation (not the Apache server). Chapter 18 shows you how to configure multiple domains.deny Allow from all AllowOverride FileInfo AuthConfig Limit Options MultiViews -Indexes SymLinksIfOwnerMatch IncludesNoExec <Limit GET POST OPTIONS> Order allow.22Harrison_ch20.conf file for the changes to take effect on the running process.

..26 www. for example. Here you could add an entry similar to: 97... Now you’ll be able to view the contents of the file my-example. . that’s what the client will use. Say. PCs and Linux workstations query the hosts file first before checking DNS.org is listed in the file.html file. you create a subdirectory named /home/www/site1/example under www.org/example/my-example.html pages in each subdirectory under your DocumentRoot directory. Apache will default to giving a listing of all the files in that subdirectory. The Windows equivalent of the Linux /etc/hosts file is named C:\ WINDOWS\system32\drivers\etc\hosts.html in this subdirectory if you point your browser to: http://www. so if a value for www.org Do not remove the localhost entry in this file.org/example www. because if one isn’t found.my-web- Apache lists all the contents of the files in the example directory if it can’t find the index. Note When setting up a yum server it’s best to enable directory listings for the RPM subdirectories.my-web-site.qxd 2/25/05 10:11 AM Page 326 326 The Apache Web Server Chapter 20 By default..html site. You need to open and edit it with a text editor. Options MultiViews -Indexes SymLinksIfOwnerMatch IncludesNoExec Remember to restart Apache after the changes. Users attempting to access the nonexistent index page will now get a “403 Access Denied” message.my-web-site. If curious surfers decide to see what the index page is for they would type the link: http://www.158. You can disable the directory listing by using a Indexes option in the <Directory> directive for the DocumentRoot like this: <Directory “/home/www/*”> .22Harrison_ch20.my-web-site. Disabling Directory Listings Be careful to include any index... such as Notepad.my-web-site. This allows Web surfers to double-check the locations of files through their browsers. .my-web-site.org’s DocumentRoot of /home/www/site1/.253.org/example.

You can place this statement in the httpd.. transparently uncompressing the data and presenting it on the screen. You can see the missing. .0 (compatible.0.com/missing.htm Notice that this gives the same output as: http://www.htm. [root@ bigboy tmp]# .115 . USING DATA COMPRESSION ON WEB PAGES Apache also has the ability to dynamically compress static Web pages into gzip format and then send the result to the remote Web surfers’ Web browser.5.. 67.com/sendmail.qxd 2/25/05 10:11 AM Page 327 Using Data Compression on Web Pages 327 Handling Missing Pages You can tell Apache to display a predefined HTML file whenever a surfer attempts to access a non-index page that doesn’t exist. MSIE 5.linuxhomenetworking... YComp 5. After making these modifications and restarting Apache.htm” “Mozilla/4.1” 200 15190 “http://www. Compare the file sizes in this Apache log: [root@ bigboy tmp]# grep dns-static /var/log/httpd/access_log ..htm file I use by trying the nonexistent link: http://www. you will be able to verify from your /var/log/httpd/access_log file that the sizes of the transmitted HTML pages have shrunk.htm instead of a generic “404 File Not Found” message: ErrorDocument 404 /missing. .linuxhomenetworking.htm Remember to put a file with this name in each DocumentRoot directory..119.conf file and then use Location directives to specify which type of files to compress. Windows NT 4. This can significantly reduce bandwidth charges if you are paying for Internet access by the megabyte.22Harrison_ch20.. AT&T CSM6..6)” .[15/Feb/2003:23:06:51 -0800] “GET /dns-static.htm HTTP/1.linuxhomenetworking.25.0.com/bogus-file. which will make Apache display the contents of missing.. Most current Web browsers support this format.2.0. First you need to load Apache version 2’s deflate module in your httpd.conf file.

.conf loads the compression module mod_deflate by default. Note Fedora’s version of httpd.htm -rw-r--r-1 user group 78350 Feb 15 00:53 /home/www/ccie/dns-static.x has some problems. but it is fine BrowserMatch \bMSIE !no-gzip !gzip-only-text/html # Don’t compress images SetEnvIfNoCase Request_URI \ \. however.08 have some more problems BrowserMatch ^Mozilla/4\. LoadModule deflate_module modules/mod_deflate.190 bytes—that’s almost 80% compression.22Harrison_ch20.(?:gif|jpe?g|png)$ no-gzip dont-vary # Make sure proxies don’t deliver the wrong content Header append Vary User-Agent env=!dont-vary </Location> .conf file to activate the compression of static pages.qxd 2/25/05 10:11 AM Page 328 328 The Apache Web Server Chapter 20 and the corresponding directory listing: [root@ bigboy tmp]# ll /web-dir/dns-static. The location statements are required.htm [root@bigboy tmp]# As you can see. 78.so <Location /> # Insert filter SetOutputFilter DEFLATE # Netscape 4..0[678] no-gzip # MSIE masquerades as Netscape.350 bytes shrunk to 15.06-4. Compression Configuration Example You can insert these statements just before your virtual hosting section of your httpd. BrowserMatch ^Mozilla/4 gzip-only-text/html # Netscape 4. Remember to restart Apache when you do. This means that the LoadModule line (the first line of the example snippet) is not required for Fedora.

com of www.org ServerAlias bigboy.1.1.1. For example.253.100 when visiting the same site.another-web-site. If you are on the inside.100.168. then you may encounter problems when trying to access www. www.158.168. Fortunately Apache allows you to specify multiple IP addresses in the <VirtualHost> statements to help you overcome this problem.X network.100 NameVirtualHost 97.168. Users on the Internet need to access IP address 97.158.26 via NAT/masquerading.253.org </VirtualHost> Step 2: Configure DNS Views You now need to fix the DNS problem that NAT creates. Linux Web server Bigboy has an internal IP address of 192.1.158.1. There is a two-part solution to this problem: Step 1: Configure Virtual Hosting on Multiple IPs You can configure Apache to serve the correct content when accessing www. but the firewall presents it to the world with an external IP address of 97.org from the outside.org and users on your home network need to access IP address 192. 192. Chapter 18 explains how to do this in detail.26 when visiting www.253. and also when accessing the specific IP address 192. Here is an example: NameVirtualHost 192.253.mysite.my-web-site.100 97.168.100 from the inside.26 <VirtualHost 192. firewalls frequently don’t allow access from their protected network to IP addresses that they masquerade on the outside.158.qxd 2/25/05 10:11 AM Page 329 Apache Running on a Server Behind a NAT Firewall 329 APACHE RUNNING ON A SERVER BEHIND A NAT FIREWALL If your Web server is behind a NAT firewall and you are logged on a machine behind the firewall as well.253.168.mysite.org.158. Because of NAT (network address translation). . You can configure your DNS server to use views which makes your DNS server give different results depending on the source IP address of the Web surfer’s PC doing the query.com or www.22Harrison_ch20.26> DocumentRoot /www/server1 ServerName www.my-web-site.26.168. you may find it impossible to hit URLs that resolve in DNS to 97.1.another-web-site.another-web-site.

htpasswd AuthGroupFile /dev/null AuthName EnterPassword AuthType Basic require user peter Remember this password protects the directory and all its subdirectories. Use Apache’s htpasswd password utility to create username/password combinations independent of your system login password for Web page access. Make the . then you can edit your PC’s hosts file as a quick and dirty temporary solution to the problem. switch on the command line.htpasswd paul New password: Re-type new password: Adding password for user paul [root@bigboy tmp]# 2.qxd 2/25/05 10:11 AM Page 330 330 The Apache Web Server Chapter 20 Tip If you have to rely on someone else to do the DNS change. but require passwords for directories or pages that show MRTG or Webalizer data. You have to specify the location of the password file. PROTECTING WEB PAGE DIRECTORIES WITH PASSWORDS You can password protect content in both the main and subdirectories of your DocumentRoot fairly easily. Here is an example for a first user named peter and a second named paul: [root@bigboy tmp]# htpasswd -c /etc/httpd/conf/. I recommend placing the file in your /etc/httpd/conf directory.22Harrison_ch20. away from the DocumentRoot tree where Web users could possibly view it. I know people who allow normal access to their regular Web pages. or create. The require user statement tells Apache that only user peter in the . and if it doesn’t yet exist.htpasswd file readable by all users: [root@bigboy tmp]# chmod 644 /etc/httpd/conf/.htpasswd peter New password: Re-type new password: Adding password for user peter [root@bigboy tmp]# [root@bigboy tmp]# htpasswd /etc/httpd/conf/. The AuthUserFile tells Apache to use the .htaccess file in the directory to which you want password control with these entries: AuthUserFile /etc/httpd/conf/. This example shows how to password protect the /home/www directory: 1. Create a . you have to include a -c. Remember that this will fix the problem on your PC alone.htpasswd file should .htpasswd file.htpasswd 3.

THE /ETC/HTTPD/CONF.htaccess file in the directory /home/www: [root@bigboy tmp]# chmod 644 /home/www/. Restart Apache. all directories below /var/www/ require password authorization: <Directory /home/www/*> AllowOverride AuthConfig </Directory> 6. in which a Web server has to host many Web sites.htpasswd users to have access.conf file.conf file has an AllowOverride statement in a <Directory> directive for any directory in the tree above /home/www. 4. If it is shared.conf file.d directory don’t have to have any special names. If each site has a dedicated IP address. .d directory file. 2. If you want all .conf file every time Apache is restarted. To do this correctly: 1.d Directory 331 have access. 3. Remove the corresponding directives from the httpd. Make sure your /etc/httpd/conf/http.22Harrison_ch20. Restart Apache. it’ll need to remain in the main httpd. The files located in the /etc/httpd/conf.158. and test.conf file. you can create one configuration file per Web site each with its own set of <VirtualHost> and <Directory> containers.htaccess 5. replace this line with require valid-user.D DIRECTORY Files in the /etc/httpd/conf.qxd 2/25/05 10:11 AM Page 331 The /etc/httpd/conf. when you try accessing the Web site. In complicated configurations. AuthType Basic instructs Apache to accept basic unencrypted passwords from the remote users’ Web browser. then place the NameVirtualHost statements in the corresponding /etc/httpd/conf.253. in case you make a mistake. Create the files located in this directory that contain the Apache required <VirtualHost> and <Directory> containers and directives. Make sure that you have a <VirtualHost> directive that defines access to /home/www or another directory higher up in the tree: <VirtualHost *> ServerName 97.conf file. In this example below. 4. and you don’t have to refer to them in the httpd. Now. This can make Web site management much simpler. 5.d directory are read and automatically appended to the configuration in the httpd.26 DocumentRoot /home/www </VirtualHost> 7. Back up your httpd. you’ll be prompted for a password. Set the correct file protections on your new .

you’ll find errors in the /var/log/httpd/error_log file during normal operation or displayed on the screen when Apache starts up. It is best to test this from both inside your network and from the Internet. Nov 21 20:41:23 bigboy kernel: audit(1101098483. A sure sign of problems related to security context are “avc: denied” messages in your /var/log/messages log file. or port forwarding rules to your Web server. Most of the errors you’ll encounter will probably be related to incompatible syntax in the <VirtualHosts> statement caused by typing errors. . If you can connect on port 80 but no pages are being served. not the Web server software itself. Other sources of failure include Apache not being started at all. then the problem is usually due to a bad Web application. The second cause is usually related to an incorrect security context for the Web page’s file. or networkrelated failures. Please refer to the “General Configuration Steps” section for further details. “Simple Network Troubleshooting. the server being down. Lack of connectivity could be caused by a firewall with incorrect permit. Troubleshooting with TELNET is covered in Chapter 4.html file in your Web site’s DocumentRoot directory.” Browser 403 Forbidden Messages Browser 403 Forbidden messages are usually caused by file permissions and security context issues.22Harrison_ch20. there are two main causes. NAT.html dev=hda5 ino=12 scontext=root:system_r:httpd_t tcontext=root:object_r:home_root_t tclass=file Only the Default Apache Page Appears When only the default Apache page appears.qxd 2/25/05 10:11 AM Page 332 332 The Apache Web Server Chapter 20 TROUBLESHOOTING APACHE Troubleshooting a basic Apache configuration is fairly straightforward. Testing Basic HTTP Connectivity The very first step is to determine whether your Web server is accessible on TCP port 80 (HTTP). The first is the lack of an index.897:0): avc: denied { getattr } for pid=1377 exe=/usr/sbin/httpd path=/home/www/index. Please refer to the “General Configuration Steps” section for further details.

For the simple virtual hosting example above. it would be easiest to: 1.X.1 for ServerName Starting httpd: [Wed Feb 04 21:18:16 2004] [error] (EAI 2)Name or service not known: Failed to resolve server name for 192.rpmnew.conf. . Save the old httpd.rpmnew file.X default configuration file is stored in /etc/httpd/conf/httpd. and VirtualHost containers from the old file and place them in the new httpd.22Harrison_ch20.conf Files When Upgrading Your old configuration files will be incompatible when upgrading from Apache version 1. httpd.3 to Apache 2.1.conf-version-1. and name it httpd. or else you’ll get an error similar to these when starting httpd: Starting httpd: httpd: Could not determine the server’s fully qualified domain name.or specify an explicit ServerName You can avoid this by adding a default generic ServerName directive at the top of the httpd.conf. Server Name Errors All ServerName directives must list a domain that is resolvable in DNS. Copy the ServerName.conf file with another name. NameVirtualHost. using 127.name:80: #ServerName new.x for example. 2.host.name:80 ServerName localhost The Apache Status Log Files The /var/log/httpd/access_log file is updated after every HTTP query and is a good source of general purpose information about your Web site.conf.0.3 lists the layout.conf file that references localhost instead of the default new.16. There is a fixed formatting style with each entry separated by spaces or quotation marks.host.100 (check DNS) -.0. Restart Apache.rpmnew file. 3.conf. Copy the httpd. The new version 2.qxd 2/25/05 10:11 AM Page 333 Troubleshooting Apache 333 Incompatible /etc/httpd/conf/http. Table 20.

Unauthorized access. Successful request.htm using Microsoft Internet Explorer version 5.119. The amount of data sent was 15190 bytes and the surfer got to the site by clicking on the link http://www.119. Often occurs when the Web page file is owned by user root. Someone entered an incorrect username or password on a password-protected page.115 .0.0. not the operation of the Apache server.25.qxd 2/25/05 10:11 AM Page 334 334 The Apache Web Server Chapter 20 Table 20. Table 20. YComp 5.25..0. Forbidden. AT&T CSM6. For example.com/sendmail.22Harrison_ch20. This could be caused by incorrect URL links in other pages on your site.5.2. Windows NT 4.htm” “Mozilla/4.com/sendmail. 67.115 on February 15 looked at the Web page /dns-static.4 HTTP Status Codes Description Successful request. even though it has universal read access. HTTP Code 200 304 401 403 .4 has some of the more common examples. but the Web page requested hasn’t been modified since the current version in the remote Web browser’s cache.6)” The HTTP status code can provide some insight into the types of operations surfers are trying to attempt and may help to isolate problems with your pages.linuxhomenetworking.htm returning a successful 200 status code. you can determine that someone at IP address 67. 404 errors are generated when someone tries to access a Web page that doesn’t exist anymore.htm HTTP/1. MSIE 5. linuxhomenetworking.[15/Feb/2003:23:06:51 -0800] “GET /dns-static. Frequently occurs when a surfer is browsing back and forth on a site.5.1” 200 15190 “http://www.3 Apache Log File Format Description IP address of the remote Web surfer Time stamp HTTP query including the Web page served HTTP result code The amount of data in bytes sent to the remote Web browser The Web page that contained the link to the page served The version of the Web browser used to get the page Separator Spaces Square Brackets [] Quotes “” Spaces Spaces Quotes “” Quotes “” Field Number 1 2 3 4 5 6 7 Upon examining the entry.0 (compatible. This means the Web page will not be sent to the remote browser. Table 20. File permissions prevent Apache from reading the file. it will just use its cached version instead.

CONCLUSION Web sites both personal and commercial can be very rewarding exercises as they share your interests with the world and allow you to meet new people with whom to develop friendships or transact business. which can be very useful. Unfortunately. .22Harrison_ch20. even the best Web sites can be impersonal as they frequently only provide information that the designer expects the visitor to need. The file will also include Apache startup errors. The Apache Error Log Files The /var/log/httpd/error_log file is a good source for error information. Check your error_log file for further details on the script’s error message. The /var/log/httpd/error_log file also is the location where CGI script errors are written. has the advantage of being able to relay documents and other information without interrupting the addressee. Frequently generated by CGI scripts that fail from bad syntax. Chapter 21. Unlike the /var/log/httpd/access_log file. “Configuring Linux Mail Servers. This allows them to schedule a response when they are better prepared to answer.” explains how to configure a Linux e-mail server to reduce spam and provide personalized addresses across multiple domains. Page requested doesn’t exist. E-mail. there is no standardized formatting. although ancient in comparison to newer personalized interactive Internet technologies. Many times CGI scripts fail with a blank screen on your browser. Internal server error. No Web site should be without one. such as IP telephony and instant messaging.qxd 2/25/05 10:11 AM Page 335 Conclusion 335 HTTP Code 404 500 Description Not found. a valuable quality when replies need to be complex. the /var/log/httpd/error_log file most likely lists the cause of the problem. Typical errors that you’ll find here are HTTP queries for files that don’t exist or forbidden requests for directory listings.

22Harrison_ch20.qxd 2/25/05 10:11 AM Page 336 .

sendmail can handle both incoming and outgoing mail for your domain. but if you are running a business.23Harrison_ch21. are also gaining popularity with Linux. CONFIGURING SENDMAIL One of the tasks in setting up DNS for your domain (my-web-site. In a home environment. This chapter will show you how to use sendmail to create a mail server that will relay your mail to a remote user’s mailbox or incoming mail to a local mailbox.org) is to use the MX record in the configuration zone file to state the hostname of the server that will handle the mail for the domain. The most popular Unix mail transport agent is sendmail.qxd 2/25/05 10:38 AM Page 337 C H A P T E R 21 Configuring Linux Mail Servers IN THIS CHAPTER ☞ ☞ ☞ ☞ Configuring sendmail Fighting Spam Configuring Your POP Mail Server Conclusion E-mail is an important part of any Web site you create. but others. Take a closer look. You’ll also learn how to retrieve and send mail via your mail server using a mail client such as Outlook Express or Evolution. 337 . The steps used to convert a Linux box into a sendmail mail server will be explained here. then a dedicated mail server will probably be required. How sendmail Works As stated before. a free Web-based e-mail service may be sufficient. such as postfix and qmail.

Users retrieve their mail from the mail server using client software. Mail sent to each of these users (username@my-web-site. then your mail server could be commandeered to relay spam.org) eventually arrives at your mail server and sendmail then processes it and deposits it in the mailbox file of the user’s Linux account. Using a series of highly complicated regular expressions listed in the /etc/mail/sendmail. a much simpler file named /etc/sendmail.cf file.qxd 2/25/05 10:38 AM Page 338 338 Configuring Linux Mail Servers Chapter 21 Incoming Mail Usually each user in your home has a regular Linux account on your mail server. sendmail inspects this header and then acts accordingly. These are then interpreted by a number of macro routines to create the sendmail. PC and Linux workstation users configure their e-mail software to make the mail server their outbound SMTP mail server. Note If mail relaying is not configured properly.com domain. One of the main advantages of mail relaying is that when a PC user A sends mail to user B on the Internet. such as Evolution. After editing sendmail. Mail isn’t actually sent directly to the user’s PC. according to the desires of the systems administrator. the PC of user A can delegate the SMTP processing to the mail server. If the mail is destined for a local user in the mysite. then sendmail places the message in that person’s mailbox so that they can retrieve it using one of the methods above. and sometimes modified. or a GUI client.cf file. sendmail first uses DNS to get the MX record for the other domain. Simple sendmail security will be covered later. that supports either the POP or IMAP mail retrieval protocols.23Harrison_ch21. Linux users logged into the mail server can read their mail directly using a text-based client. Linux workstation users can use the same programs to access their mail remotely. such as Microsoft’s Outlook or Outlook Express.mc. . If the mail is being sent to another domain. It then attempts to relay the mail to the appropriate destination mail server using the Simple Mail Transport Protocol (SMTP). In recognition of the complexity of the /etc/mail/sendmail. you must always run the macros and restart sendmail for the changes to take effect. Outgoing Mail The process is different when sending mail via the mail server. such as mail. sendmail Macros When mail passes through a sendmail server the mail routing information in its header is analyzed.mc was created. and it contains more understandable instructions for systems administrators to use.cf file.

mc directive starts with a keyword.mc file. in this case sendmail. followed by a subdirective and in some cases arguments. in which the macro may be found and the subdirective is usually the name of the macro file itself. The macros in the /usr/share/sendmail-cf/m4 directory deal with these. and redirect the output. Notice that sendmail is sensitive to the quotation marks used in the m4 macro directives.mc file later in the chapter.mc. you can then execute the make command while in the /etc/mail directory to regenerate the new sendmail. or OSTYPE. Once you finish editing the sendmail. # m4 /etc/mail/sendmail. So in the example.`hash -o /etc/mail/virtusertable. such as define for the definition of certain sendmail variables and MASQUERADE_DOMAIN.23Harrison_ch21.db’)dnl The keywords usually define a subdirectory of /usr/share/sendmail-cf. A typical example is: FEATURE(`virtusertable’.m4. .cf I’ll discuss many of the features of the sendmail. you just specify the name of the macro file as the argument. FEATURE. to the sendmail. The m4 usage is simple. such as DOMAIN.cf file with the > redirector symbol.cf file. [root@bigboy mail]# The make command actually generates the sendmail.db’ is being passed to it. They open with a grave mark and end with a single quote: FEATURE(`masquerade_envelope’)dnl Some keywords. have no corresponding directories with matching macro files. and the instruction `\ hash -o /etc/ mail/virtusertable. the macro name is /usr/share/ sendmail-cf/feature/virtusertable. [root@bigboy tmp]# cd /etc/mail [root@bigboy mail]# make make If there have been no changes to the files in /etc/mail since the last time was run. then you’ll get an error like this: [root@bigboy mail]# make make: Nothing to be done for `all’.qxd 2/25/05 10:38 AM Page 339 Configuring sendmail 339 Each sendmail.cf file using the m4 command.mc > /etc/mail/sendmail. which would normally go to the screen.

which creates a new sendmail. remember that the filename usually starts with the software package name by a version number. and m4 software RPMs are installed.mc file and compiles supporting configuration files in the /etc/mail directory according to the instructions in the file /etc/mail/Makefile.1.” will tell you how. stop. [root@bigboy tmp]# pgrep sendmail You should get a response of plain old process ID numbers. as in sendmail8.qxd 2/25/05 10:38 AM Page 340 340 Configuring Linux Mail Servers Chapter 21 Installing sendmail Most Red Hat and Fedora Linux software products are available in the RPM format.10-1. use: [root@bigboy tmp]# service sendmail start [root@bigboy tmp]# service sendmail stop [root@bigboy tmp]# service sendmail restart Remember to restart the sendmail process every time you make a change to the configuration files for the changes to take effect on the running process.i386. How to Restart sendmail After Editing Your Configuration Files In this chapter.1. the . You can also test whether the sendmail process is running with the pgrep command.) When searching for the RPMs. Starting sendmail You can use the chkconfig command to get sendmail configured to start at boot: [root@bigboy tmp]# chkconfig sendmail on To start.cf file from sendmail. You will need to make sure that the sendmail.12. (Chapter 6. you’ll see that sendmail uses a variety of configuration files that require different treatments for their commands to take effect.23Harrison_ch21. “Installing RPM Software. This little script encapsulates all the required post configuration steps: #!/bin/bash cd /etc/mail make newaliases /etc/init. and restart sendmail after booting.d/sendmail restart It first runs the make command.rpm. sendmail-cf.

mc file doesn’t use this character for commenting. Configuration of the sendmail.mc is a more user friendly The /etc/mail/sendmail. but I included it so that you don’t forget. you can define most of sendmail’s configuration parameters in the /etc/mail/sendmail. it reads sendmail. and /etc/mail/sendmail. The /etc/sendmail.23Harrison_ch21.cf file is located in different directories depending on the version of Red Hat you use. The sendmail.cf file is used for versions up to 7.Addr=127.0.0 and higher and Fedora Core. Note the file sendmail. Fortunately.cf. How to Put Comments in sendmail.mc File As discussed earlier. and then restarts sendmail.0. This may not be a good idea in a production system. Use this command to make the script executable: chmod 700 filename You’ll need to run the script each time you change any of the sendmail configuration files. which is then used by the m4 macros to create the /etc/mail/sendmail.Addr=127.1. configuration file and really is much easier to fool around with without getting burned.cf file. Name=MTA’) dnl # DAEMON_OPTIONS(`Port=smtp.1.qxd 2/25/05 10:38 AM Page 341 Configuring sendmail 341 It then generates new e-mail aliases with the newaliases command. The line in the script that restarts sendmail is only needed if you have made changes to the /etc/mail/sendmail.cf is used for versions 8. When sendmail starts.mc file is much simpler than configuration of sendmail.0.mc file. (this will be covered later). Name=MTA’) .3. The sendmail.0.mc configuration file: These statements are disabled by dnl commenting: dnl DAEMON_OPTIONS(`Port=smtp. Here are some valid examples of comments used with the sendmail. but instead uses the string “dnl”.mc file.cf for its configuration.mc In most Linux configuration files a # symbol is used at the beginning of a line to convert it into a comment line or to deactivate any commands that may reside on that line. but it is still often viewed as an intimidating task with its series of structured directive statements that get the job done. in most cases you won’t have to edit this file very often.

The program also expects the files used by the mail server’s DNS client to be configured correctly.1. See either Chapter 18.0.1. then bigboy.1.conf File The sendmail program expects DNS to be configured correctly on the DNS server. You can test this using the host command: [root@smallfry tmp]# host bigboy.org nameserver 192.0. “Dynamic DNS.1.” or Chapter 19. Configure Your Mail Server’s Name In DNS You first need to make sure that your mail server’s name resolves in DNS correctly.1.168. Configure the /etc/resolv.org and the IP address of the DNS server is 192.100 [root@smallfry tmp]# You will need to fix your DNS server’s entries if the resolution isn’t correct. there must be directives that look like this: domain my-web-site. “Configuring DNS.my-site.Addr=127.100. if the mail server is handling mail for my-web-site.Addr=127.my-web-site.168.org.0. Name=MTA’) Configuring DNS for sendmail Remember that you will never receive mail unless you have configured DNS for your domain to make your new Linux box mail server the target of the DNS domain’s MX record.com bigboy.conf file in which there must be a domain directive that matches one of the domains the mail server is expected to handle mail for. The first one is the /etc/resolv.168.org must correctly resolve to the IP address of one of the mail server’s interfaces. sendmail expects a nameserver directive that points to the IP address of the DNS server the mail server should use to get its DNS information.23Harrison_ch21.” for details on how to do this. For example. For example. if your mail server’s name is Bigboy and you intend for it to mostly handle mail for the domain my-web-site.100 .my-website. The MX record for your domain must point to the IP address of the mail server. Name=MTA’) This statement is active: DAEMON_OPTIONS(`Port=smtp.0. Finally.qxd 2/25/05 10:38 AM Page 342 342 Configuring Linux Mail Servers Chapter 21 This statement is incorrectly disabled: # DAEMON_OPTIONS(`Port=smtp.org has address 192.

localdomain. To . the mail server has been set to mail.1 entry in /etc/hosts doesn’t also include localhost and localhost.`mail.1 must always be followed by the fully qualified domain name (FQDN) of the server.org. you can add any other aliases your host may have to the end of the line. This is configured in the sendmail.qxd 2/25/05 10:38 AM Page 343 Configuring sendmail 343 An incorrectly configured resolv. If the sendmail server is a Linux server. it would be bigboy. Then you must have an entry for localhost and localhost. To do this. you need to process the sendmail.myweb-site.0.mc file: WARNING: local host name (smallfry) is not qualified.org’) If you don’t have a mail server on your network. define(`SMART_HOST’. Here is a brief example of the first line you should expect to see in it: 127.0. sendmail therefore uses the loopback address to send mail between users on the same Linux server. Linux does not function properly if the 127.1.org localhost.my-web-site. How to Configure Linux sendmail Clients All Linux mail clients in your home or company need to know which server is the mail server.0. fix $j in config file The /etc/hosts File The /etc/hosts file also is used by DNS clients and also needs to be correctly configured. Converting From a Mail Client to a Mail Server All Linux systems have a virtual loopback interface that lives only in memory with an IP address of 127. As mail must be sent to a target IP address even when there is no NIC in the box.localdomain.mc file by setting the SMART_HOST statement to include the mail server. then the /etc/hosts file will also have to be correctly configured.23Harrison_ch21.localdomain localhost bigboy The entry for 127. In the example below. run the restarting script from earlier in the chapter.0.1 bigboy.conf file can lead to errors when running the m4 command to process the information in your sendmail. the mail server for the my-web-site.my-web-site.0.mc file and restart sendmail.0.org domain. In the case above.org.my-web-site.0. Finally. Once this is done.0. you can either create one or use the one offered by your ISP.

0. you use netstat and grep for 25 to see a default configuration listening only on IP address 127.1. DAEMON_OPTIONS(`Port=smtp. as well: 1. Name=MTA’) We strongly recommend to comment this one out if you want to protect yourself from spam. you should comment out the daemon_options line in the /etc/mail/sendmail.1:25 0.0.0:* LISTEN [root@bigboy tmp]# 2.0.0. Bigboy the mail server does not accept e-mail relayed from any of the other PCs on your network if they are not in DNS. then you can uncomment the localhost DAEMON_OPTIONS entry and add another one for the IP address of the NIC on which you wish to accept SMTP traffic.Addr=127. If sendmail is listening on the loopback interface only. Edit sendmail. .0.0. It is also good practice to take precautions against spam by not accepting mail from domains that don’t exist by commenting out the accept_ unresolvable_domains feature too. FEATURE(`accept_unresolvable_domains’)dnl FEATURE(`relay_based_on_MX’)dnl Note You need to be careful with the accept_unresolvable_names feature.0. Comment this out if you want to accept email over the network.1 (loopback): [root@bigboy tmp]# netstat -an | grep :25 | grep tcp tcp 0 0 127. Determine which NICs sendmail is running on. In the sample network. Chapter 18 shows how to create your own internal domain just for this purpose.1 and not on any other network devices.0. See the sixth and next to last lines in the example: dnl dnl dnl dnl dnl dnl dnl … … … dnl dnl dnl dnl dnl dnl dnl dnl This changes sendmail to only listen on the loopback device 127. However. Because sendmail listens on TCP port 25. sendmail needs to be configured to listen for messages on NIC interfaces.mc file with dnl statements. You can see the interfaces on which sendmail is listening with the netstat command. Note If your server has multiple NICs and you want it to listen to only one of them.23Harrison_ch21. and not a mail client.0.0.mc to make sendmail listen on all interfaces. the laptop and users on computers that do not have 24x7 DNS do need this.qxd 2/25/05 10:38 AM Page 344 344 Configuring Linux Mail Servers Chapter 21 become a mail server.

`mail.0:25 0.mc.org and not destined for this mail server will be forwarded: another-web-site.org’) 4.mc file. and restart sendmail. you can do this with the restart script from the beginning of the chapter. [root@bigboy tmp]# netstat -an | grep :25 | grep tcp tcp 0 0 0. Here is a sample: dnl ***** Customised section 1 start ***** dnl dnl FEATURE(delay_checks)dnl FEATURE(masquerade_envelope)dnl FEATURE(allmasquerade)dnl FEATURE(masquerade_entire_domain)dnl dnl dnl dnl ***** Customised section 1 end ***** The /etc/mail/relay-domains File The /etc/mail/relay-domains file is used to determine domains from which it will relay mail. Regenerate the sendmail.0.mc file can seem jumbled. A General Guide to Using the sendmail. The contents of the relay-domains file should be limited to those domains that can be trusted not to originate spam. this file does not exist in a standard Red Hat/Fedora install. The following sections will show you how to define what type of mail it should handle and the various ways this mail can be processed. The mail server doesn’t need a SMART_HOST entry in its sendmail. To make it less cluttered. Sometimes sendmail will archive this file when you do a version upgrade. Comment this out with a dnl at the beginning.0. dnl define(`SMART_HOST’.mc File The sendmail.my-web-site. all mail sent from another-web-site. and the second section is at the very bottom. Make sure sendmail is listening on all interfaces (0. The first section is near the top where the FEATURE statements usually are. I usually create two easily identifiable sections in it with all the custom commands I’ve ever added.qxd 2/25/05 10:38 AM Page 345 Configuring sendmail 345 3.23Harrison_ch21.0. By default.0.0.org .0:* LISTEN [root@bigboy tmp]# You have now completed the first phase of converting your Linux server into a sendmail server by enabling it to listen to SMTP traffic on its interfaces. Again.cf file.0. Having easily identifiable modifications in the file will make post upgrade reconfiguration much easier. 5. Comment out the SMART_HOST entry in sendmal.0). In this case.

168.168. my experience has been that control on a per e-mail address basis is much more intuitive via the /etc/mail/virtusertable file. If you delete /etc/mail/relay-domains. only spam flowing through you. The sample file that follows allows relaying for only the server itself (127. The first lists IP addresses and domains from which the mail is coming or going.1.1. the mail server will relay mail only for those PCs on your network that have their e-mail clients configured to use the mail server as their outgoing SMTP mail server.org RELAY RELAY RELAY RELAY RELAY RELAY RELAY . localhost). That is to say.23Harrison_ch21. and DISCARD. The /etc/mail/access file has two columns. The /etc/mail/access file has more capabilities. sendmail assumes it could be either and tries to match both. everyone on your 192. and everyone passing e-mail through the mail server from servers belonging to my-web-site.1. such as restricting relaying by IP address or network range and is more commonly used. There is no third column to state whether the IP address or domain is the source or destination of the mail.168.17 192.X network. and source domains can be spoofed by spam e-mail servers. sendmail will reject. All other attempted relayed mail that doesn’t match any of the entries in the /etc/mail/access file.16 192.) If you don’t take the precaution of using this feature.0. you may find your server being used to relay mail for spam e-mail sites. Keywords include RELAY. Configuring the /etc/mail/access file will not stop spam coming to you. The /etc/mail/access File You can make sure that only trusted PCs on your network have the ability to relay mail via your mail server by using the /etc/mail/access file.168.qxd 2/25/05 10:38 AM Page 346 346 Configuring Linux Mail Servers Chapter 21 One disadvantage of this file is that it controls mail based on the source domain only.localdomain localhost 127.2 my-web-site. two client PCs on your home 192.2. OK (not ACCEPT). then relay access is fully determined by the /etc/mail/access file.X network.168. Despite this. Be sure to run the restart sendmail script from the beginning of the chapter for these changes to take effect.org only if its IP address can be found in a DNS reverse zone file: localhost. The second lists the type of action to be taken when mail from these sources or destinations is received.org. (In Outlook Express.0.1.1 192. REJECT.0. Remember that a server will be considered a part of my-web-site.0. you set this using: Tools>Accounts>Properties>Servers.

org another-web-site.org DNS In this case.23Harrison_ch21. (Remember each is important. This file has a list of hostnames and domains for which sendmail accepts responsibility. Remember that the relay security features of this file may not work if you don’t have a correctly configured /etc/hosts file. Here is an example. and the second column lists the local user’s mail box. The /etc/mail/virtusertable file sendmail checks the aliases The /etc/mail/virtusertable file contains a set of simple instructions on what to do with received mail. Which User Should Really Receive the Mail? After checking the contents of the virtusertable. If there is no match in the virtusertable file. or a mailing list entry in the /etc/aliases file to which the e-mail should be forwarded. Here are the commands you need: [root@bigboy tmp]# cd /etc/mail [root@bigboy mail]# make The sendmail restart script configured at the beginning of the chapter does this for you too.org.org. remember to modify the MX record of the another-webzone file to point to my-web-site.my-web-site. files to determine the ultimate recipient of mail. a remote e-mail address. MX 10 mail. The first column lists the target e-mail address. It uses the /etc/mail/local-host-names file to do this.org and another-site then the file would look like this: my-web-site. Primary Mail Exchanger for another-web-site.org.) . sendmail checks for the full e-mail address in the /etc/aliases file: . if this mail server was to accept mail for the domains my-web-site.org another-web-site. it needs a way of determining whether it is responsible for the mail it receives.qxd 2/25/05 10:38 AM Page 347 Configuring sendmail 347 You’ll then have to convert this text file into a sendmail readable database file named /etc/mail/access.db. For example.org site. The /etc/mail/local-host-names File When sendmail receives mail.

mail sent to: webmaster@another-web-site. all other mail to another-web-site. lp. and when it is all finished.org @another-web-site. named. The /etc/aliases File You can think of the /etc/aliases file as a mailing list file. sendmail notes the nickname entry in the second column.org. It then searches for the nickname again in the first column to see if the recipient isn’t on yet another mailing list. daemon. and so on by system processes will all be sent to user (or mailing list) root. you have: They’re in your all-purpose sendmail restart script. If sendmail doesn’t find a duplicate. apache.com ☞ paul and finance at my-web-site.org goes to local user (or mailing list) paul ☞ All other users at my-web-site.org goes to local user marc ☞ sales at my-web-site.qxd 2/25/05 10:38 AM Page 348 348 Configuring Linux Mail Servers Chapter 21 webmaster@another-web-site.org paul@my-web-site. and the second column has the members of the mailing list separated by commas. you can see that mail sent to users bin. To start. In this case.” After editing the /etc/mail/virtusertable file. . The first column has the mailing list name (sometimes called a virtual mailbox). If it finds a match in the first column. In the example that follows.org sales@my-web-site.org goes to local user (or mailing list) webmasters.org @my-web-site. they all get a copy of the e-mail message.org goes to the sales department at my-othersite. then sendmail goes through the process all over again to determine if any of the members is on yet another list. If there is no match. root is actually an alias for a mailing list consisting of user marc and webmaster@my-web-site. it assumes the recipient is a regular user on the local server and deposits the mail in their mailbox.db with two commands: [root@bigboy tmp]# cd /etc/mail [root@bigboy mail]# make If these lines look like you’ve seen them before.org receive a bounce back message stating “User unknown. sendmail searches the first column of the file for a match. you have to convert it into a sendmail-readable database file named /etc/mail/virtusertable.org finance@my-web-site. shutdown.23Harrison_ch21.org paul paul error:nouser User unknown In this example. then sendmail assumes the recipient is a regular user on the local server and deposits the mail in their mailbox. If the recipient is a mailing list.org webmasters marc sales@another-web-site.

and mary: # Directors of my SOHO company directors: peter. mailer-daemon: postmaster postmaster: root # General redirections for pseudo accounts.org goes to users grandma. . bin: root daemon: root . [root@bigboy tmp]# newaliases # Basic system aliases -.org goes to users peter.. abuse: root # trap decode to catch security attacks decode: root # Person who should get root’s mail root: marc.paul. and grandma. In this simple mailing list example. brother.org.. here are a few more list examples for your /etc/aliases file. you’ll have to convert it into a sendmail readable database file named /etc/aliases. You may want to delete the comment and change user marc to another user.org Notice that there are no spaces between the mailing list entries for root: You will get errors if you add spaces. The newaliases command is used to do this.mary ☞ Mail sent to sister: # My family family: family@my-web-site.. mail sent to root actually goes to user account marc and webmaster@my-web-site.webmaster@my-web-site. ☞ Mail to directors@my-web-site.these MUST be present.23Harrison_ch21.db.brother. paul. Because aliases can be very useful. Also.sister ☞ Mail sent to admin-list gets sent to all the users listed in the file /home/mailings/admin-list.. after editing this file.qxd 2/25/05 10:38 AM Page 349 Configuring sendmail 349 Note The default /etc/aliases file installed with Red Hat/Fedora has the last line of this sample commented out with a #. # My mailing list file admin-list: “:include:/home/mailings/admin-list” .

such as Outlook Express or Evolution. no matter which server originated the e-mail. mail will appear to come from mail.qxd 2/25/05 10:38 AM Page 350 350 Configuring Linux Mail Servers Chapter 21 The advantage of using mailing list files is that the admin-list file can be a file that trusted users can edit. such as Outlook Express. you may find yourself dumping legitimate mail. try making root have an alias for a user with a fully qualified domain name.com. if not. If either of these are a problem for you. based on the settings in the /etc/hosts file.) ☞ Set up masquerading to modify the domain name of all traffic originating from and passing trough your mail server.com and not ☞ Configure your e-mail client. this forces sendmail to insert the correct fields in the header. Another is that all subscriptions and unsubscriptions have to be done manually by the mailing list administrator.org. the mail has no To: in the e-mail header. user root is only needed to update the aliases file. such as majordomo.org.org sendmail Masquerading Explained If you want your mail to appear to come from user@bigboy. When sendmail sends e-mail to a local user. If you then use a mail client with a spam mail filtering rule to reject mail with no To: in the header. but you may not want your Web site to be remembered with the word “mail” in front of it. to set your e-mail address to user@mysite. (I’ll explain this in the “Configuring Your POP Mail Server” section.23Harrison_ch21. You can solve this by editing your sendmail. then you have two choices: user@mysite. In other words.mc configuration file and adding some masquerading commands and directives: . This isn’t terrible.my-web-site. One is that bounce messages from failed attempts to broadcast go to all users. you may want your mail server to handle all e-mail by assigning a consistent return address to all outgoing mail. One important note about the /etc/aliases file: By default your system uses sendmail to mail system messages to local user root. for example: # Person who should get root’s mail root: webmaster@my-web-site.mc that all outgoing mail originating on Bigboy should appear to be coming from my-web-site. Configuring Masquerading In the DNS configuration. then consider using a mailing list manager.mysite. To get around this. Despite this. you made Bigboy the mail server for the domain myweb-site. You now have to tell Bigboy in the sendmail configuration file sendmail.org. there are some problems with mail reflectors.com.

localdomain)dnl The result is that: ☞ The MASQUERADE_AS directive makes all mail originating on Bigboy appear to come from a server within the domain my-web-site. ☞ FEATURE masquerade_entire_domain makes sendmail masquerade servers named *my-web-site.’)dnl MASQUERADE_DOMAIN(localhost)dnl MASQUERADE_DOMAIN(localhost. ☞ FEATURE masquerade_envelope rewrites the e-mail envelope just as MASQUERADE_AS rewrote the header.org.org.23Harrison_ch21. the other recipient sees a cc: to an address he knows instead of one on localhost. even if the mail is sent from a user on the mail server to another user on the same mail server. ☞ always masquerades e-mail addresses.org.org and localdomain domains appear to come from the MASQUERADE_AS domain of my-web-site. FEATURE always_add_domain . Use this with caution when you are sure you have the necessary authority.com would be masqueraded. If this wasn’t selected. if MASQUERADE_AS is my-web-site.org by rewriting the e-mail header. mail from sales.org. In other words. In other words.org and my-othersite. sendmail checks the domain name associated with the IP address of the mail relay client sending the mail to help it determine whether it should do masquerading or not.localdomain. The allmasquerade statement causes all mail destined for these mailboxes to appear to be destined for users in the domain defined in the MASQUERADE_AS statement. ☞ The MASQUERADE_DOMAIN directive makes mail relayed via Bigboy from all machines in the another-web-site.com originally.org and you use allmasquerade. If you cc: yourself on an outgoing mail.org’)dnl MASQUERADE_DOMAIN(`my-web-site.org enters the correct mailbox but sendmail rewrites the To:. ☞ FEATURE allmasquerade makes sendmail rewrite both recipient addresses and sender addresses relative to the local machine.qxd 2/25/05 10:38 AM Page 351 Configuring sendmail 351 FEATURE(always_add_domain)dnl FEATURE(`masquerade_entire_domain’)dnl FEATURE(`masquerade_envelope’)dnl FEATURE(`allmasquerade’)dnl MASQUERADE_AS(`my-web-site. Using DNS.org would be masqueraded as my-web-site. then only servers named myweb-site. Note Use FEATURE allmasquerade with caution if your mail server handles e-mail for many different domains and the mailboxes for the users in these domains reside on the mail server.org as my-web-site.org and *another-web-site. then mail for peter@another-web-site.my-web-site. making the e-mail appear to be sent to peter@myste.

Using sendmail to Change the Sender’s E-mail Address Sometimes masquerading isn’t enough. perhaps you bought a program for your SOHO office that sends out notifications to your staff. For example. as spammers often do. Web-based CGI scripts tend to run as user apache and. consider the To: and From: statements as the e-mail header. To remove this restriction use the EXPOSED_USER(`root’)dnl command in /etc/mail/sendmail.qxd 2/25/05 10:38 AM Page 352 352 Configuring Linux Mail Servers Chapter 21 Masquerading is an important part of any mail server configuration as it enables systems administrators to use multiple outbound mail servers. however. not only because apache’s .mc. such as Outlook Express. Testing Masquerading The best way of testing masquerading from the Linux command line is to use the mail -v username command. The e-mail envelope contains the To: and From: used by mail servers for protocol negotiation. I have noticed that sendmail -v username ignores masquerading altogether. Other Masquerading Notes By default.23Harrison_ch21. You should also tail the /var/log/maillog file to verify that the masquerading is operating correctly and check the envelope and header of test e-mail received by test e-mail accounts. All e-mail correspondence then has a uniform e-mail address format that complies with the company’s brand marketing policies. It is the envelope’s From: that is used when e-mail rejection messages are sent between mail servers. but the program inserts its own address as sender’s address. Note E-mail clients. It is easy to fake the header. the program automatically uses the To: and From: in the header. not that of the IT person. it is detrimental to e-mail delivery. user root will not be masqueraded. Often you won’t want this. send mail as user apache too. each providing only the global domain name for a company and not the fully qualified domain name of the server itself. therefore. When you choose Reply or Reply All in Outlook Express. You can comment this out if you like with a “dnl” at the beginning of the line and running the sendmail start script. to fake the envelope. At times you may need to change not only the domain of the sender but also the username portion of the sender’s email address.

23Harrison_ch21. Your e-mails from linux-username should now appear to come from username@new-domain. actually exists as a real user. and your valid e-mail may be classified as being spam. It then looks at the /etc/mail/genericstable file for an individual e-mail linux-username username@new-domain.my-web-site. Add these statements to your feature: /etc/mail/sendmail. which you can obtain using the command: sendmail -bt -d0. then queries about the apache user will fail.qxd 2/25/05 10:38 AM Page 353 Configuring sendmail 353 e-mail address may not be suitable. Make sure the file includes your server’s canonical domain name. Troubleshooting sendmail When sendmail doesn’t appear to work correctly. If your virtusertable file allows e-mail to only predefined users. Create your /etc/mail/genericstable file. Here are a few tests and techniques you can use to fix some of the most common problems.com address from a matching domain. . Run the sendmail restart script from the beginning of the chapter and then test.org urgent-message@my-web-site.mc file to activate the FEATURE(`genericstable’. Create a /etc/mail/generics-domains file that is just a list of all the domains that should be inspected.org 3.`hash -o /etc/mail/genericstable.1 </dev/null Here is a sample /etc/mail/generics-domains file: my-web-site. you can change both the domain and username on a caseby-case basis using the genericstable feature: 1.com. With sendmail.org 4. The format of the file is: Here is an example: alert peter apache security-alert@my-web-site. First sendmail searches the /etc/mail/generics-domains file for a list of domains to reverse map. or source e-mail address.db’)dnl GENERICS_DOMAIN_FILE(`/etc/mail/generics-domains’)dnl 2.org mailer@my-web-site.org bigboy. you can test it in a number of ways.org another-web-site. but also because some anti-spam programs check to ensure that the From:.

You should get a response with a 220 status code. the cause is probably a typographical error.net [24. 3..client.org 250 2.1. If you get a “500 Command not recognized” error message along the way. Connected to mail.org 250 2. Use the hello command to tell the mail server the domain to which you belong.” on a line by itself 6.my-web-site. Failure could also be caused by the sendmail process being stopped. You should receive a message with a successful status 250 code at the beginning of the response. Using the RCPT sage is going: TO: statement. It is best to test this from both inside your network and from the Internet.my-web-site. TELNET to the mail server on port 25.org.4.org 25 Trying mail.qxd 2/25/05 10:38 AM Page 354 354 Configuring Linux Mail Servers Chapter 21 Test TCP Connectivity The very first step is to determine whether your mail server is accessible on the sendmail SMTP TCP port 25. tell the mail server to whom the test mes- RCPT TO: user@my-web-site.0 sender@another-web-site. “Simple Network Troubleshooting. Test TCP Connectivity You can also mimic a full mail session using TELNET to make sure everything is working correctly.5 user@my-web-site. end with “..my-web-site.. hello another-web-site. [root@bigboy tmp]# telnet mail. For example: .23Harrison_ch21.. Type in your text message. Type the string “subject:” then type a subject.. Recipient ok 5. or port forwarding rules to your mail server.org. ending it with a single period on the last line.org 250 mail.1.org ESMTP server ready 2. Prepare the mail server to receive data with the DATA statement: DATA 354 Enter mail.org. 220 mail. Inform the mail server from which the test message is coming with the MAIL FROM: statement: MAIL FROM:sender@another-web-site. Follow these steps carefully: 1..org Hello c-24-4-97110. Chapter 4.” covers troubleshooting with TELNET.my-web-site.comcast. Lack of connectivity could be caused by a firewall with incorrect permit. pleased to meet you. Escape character is ‘^]’. NAT.110]. Sender ok 4.my-web-site.97.org.

0 mail. SSH.my-web-site.0. [root@bigboy tmp]# Now verify that the intended recipient received the message.mc > /etc/mail/sendmail.23Harrison_ch21. Use the QUIT command to end the session. If they are not. Work in one of them and monitor the sendmail status output in the other using the command: [root@bigboy tmp]# tail -f /var/log/maillog Common Errors Due to Incomplete RPM Installation Both the newaliases and m4 commands require the sendmail-cf and m4 RPM packages. The /var/log/maillog File Because sendmail writes all its status messages in the /var/log/maillog file.cf file is out of date: sendmail 8.mc file [root@bigboy mail]# m4 /etc/mail/sendmail. and check the system logs for any mail application errors.mc:8: m4: Cannot open /usr/share/sendmailcf/m4/cf. you’ll get errors when running various sendmail related commands: ☞ Sample errors when running newaliases [root@bigboy mail]# newaliases Warning: .0 No local mailer defined .cf /etc/mail/sendmail.cf file is version 0 No local mailer defined QueueDirectory (Q) option must be set [root@bigboy mail]# ☞ Sample errors when processing the sendmail. always monitor this file whenever you are doing changes.0. .0 iA75r9si017840 Message accepted for delivery 7. QUIT 221 2.0. or console windows. These must be installed.qxd 2/25/05 10:38 AM Page 355 Configuring sendmail 355 Subject: Test Message Testing sendmail interactively .m4: No such file or directory [root@bigboy mail]# ☞ Sample errors when restarting sendmail [root@bigboy mail]# service sendmail restart Shutting down sendmail: [ OK ] Shutting down sm-client: [FAILED] Starting sendmail: 554 5.5 supports version 10. Open two TELNET.org closing connection Connection closed by foreign host.12. 250 2.

Fedora inserts the host name of the server between the 127.0. Use the sendmail program to send a sample e-mail to someone in verbose mode.another-website.org LiteMail v3.106]..org test text test text .org.1 bigboy localhost.1 (Wrong!!!) my-web-site. via esmtp. and sendmail-cf RPM By default. 220 ltmail.another-web-site...120. pleased to meet you 250 HELP .com: 127.0.0.localdomain 250-mx.221.net.23Harrison_ch21.org.org Hello [67. As mentioned.02(BFLITEMAIL4A). It will then default to sending emails in which the domain is localhost. Sat. 05 Oct 2002 06:48:44 -0400 >>> EHLO localhost. sendmail.localdomain localhost The server would therefore be open to relay all mail from any .my-web-site.com domain and would ignore the security features of the access and relaydomains files I’ll describe later. for example: [root@bigboy tmp]# sendmail -v example@another-web-site.0 QueueDirectory (Q) option must be set [FAILED] Starting sm-client: [ OK ] [root@bigboy mail]# If these errors occur. a poorly configured /etc/hosts file can make mail sent from your server to the outside world appear as if it came from users at localhost.0.. make sure your packages are installed correctly. Incorrectly Configured /etc/hosts Files m4. Connecting to mail. The /etc/hosts file is also important for configuring mail relay.qxd 2/25/05 10:38 AM Page 356 356 Configuring Linux Mail Servers Chapter 21 554 5.another-web-site.org localhost.localdomain localhost Unfortunately in this configuration.0. You can create problems if you fail to place the server name in the FDQN for 127.org.localdomain. example@another-web-site. sendmail will think that the server’s FQDN is Bigboy.0. Enter some text after issuing the command and end your message with a single period all by itself on the last line. Here sendmail thinks that the server’s FDQN was my-site and that the domain was all of .0.com or .0. which it will identify as being invalid because there is no extension at the end. such as .0.1 entry.1 and the localhost entries in /etc/hosts: 127.localdomain and not bigboy.

.qxd 2/25/05 10:38 AM Page 357 Configuring sendmail 357 >>> MAIL From:<root@localhost. You end up with an error in /var/log/maillog: Oct 16 10:20:04 bigboy sendmail[2500]: g9GHK3iQ002500: SYSERR(root): savemail: cannot save rejected email anywhere Oct 16 10:20:04 bigboy sendmail[2500]: g9GHK3iQ002500: Losing . too. such as a script based on the PERL module Mail::Audit. it is therefore an illegal Internet domain. end with “.. Sender Ok >>> RCPT To:<example@another-web-site.localdomain>.) [root@bigboy tmp]# sendmail –v root WARNING: local host name (bigboy) is not qualified.. 250 Message accepted for delivery example@another-web-site..org. PC2 sees that the mail originated from localhost. Sent (Message accepted for delivery) Closing connection to mail./qfg9GHK3iQ002500: savemail panic You may also get this error if you are using a spam prevention program.localdomain.org> 250 <example@another-web-site.localdomain> 250 <root@localhost. Consider an example: Mail sent from computer PC1 to PC2 appears to come from a user at localhost.another-web-site.localdomain is the domain that all computers use to refer to themselves..localdomain and thinks that the rejected e-mail should be sent to a user on PC2 that may not exist. fix $j in config file [root@bigboy tmp]# newaliases WARNING: local host name (bigboy) is not qualified. Recipient Ok >>> DATA 354 Enter mail. (I’ll explain the newaliases command later. sleeping for retry .org>..localdomain on PC1 and is rejected.23Harrison_ch21. Another set of telltale errors caused by the same problem can be generated when trying to send mail to a user (the example uses root) or creating a new alias database file. fix $j in config file [root@bigboy tmp]# An accompanying error in /var/log/maillog log file looks like this: Oct 16 10:23:58 bigboy sendmail[2582]: My unqualified host name (bigboy) unknown. An error in the script could cause this type of message.” on a line by itself >>> .org. The rejected e-mail is returned to localhost. >>> QUIT [root@bigboy tmp]# Localhost.

you can use your mail server to combat spam.rfc-ignorant. time consuming to delete.org’. FEATURE(`dnsbl’. if the spammer uses a public data center. `bl.shtml”’) ☞ Spamhaus: A spammer blacklist.nl DNSBL (http://proxies. `”550 5. `proxies. FEATURE(`dnsbl’.see http://www. FEATURE(`dnsbl’. Also. `’)dnl ☞ The Open Relay Database: An open mail relay list. This prevents the spammers from easily switching their servers’ IP addresses to the next available ones on their networks. In some cases.spamcop.1 ACCESS DENIED to OPEN PROXY SERVER “$&{client_name}” by easynet. but in all cases they rely heavily on complaints from spam victims. `”450 Mail from “ $`’&{client_addr} “ refused .net’. Some spam blacklists simply try to determine whether the e-mail is coming from a legitimate IP address. .nl/errors.org’.`Rejected .see http://www. FEATURE(`dnsbl’.org’.ordb.easynet. they should all be on one line: ☞ RFC-Ignorant: A valid IP address checker.23Harrison_ch21. to provide additional pressure on the spammers. Fortunately. and in some cases dangerous when they contain viruses and worms.7. `relays.net/bl.see http://spamhaus. `”550 Email rejected due to sending server misconfiguration .org/”’) ☞ Easynet: An open proxy list. It is hoped that these legitimate users will pressure the data center’s management to evict the spamming customer. it is possible that their activities could also cause the IP addresses of legitimate e-mailers to be black listed too.qxd 2/25/05 10:38 AM Page 358 358 Configuring Linux Mail Servers Chapter 21 FIGHTING SPAM Unsolicited commercial e-mail (UCE or spam) can be annoying. Rejected for bad WHOIS info on IP of your SMTP server .see http://spamcop. Some are maintained by volunteers.`sbl. The IP addresses of offenders usually remain on the list for six months to two years.blackholes.org/’)dnl Be sure to visit the URLs listed to learn more about the individual services.`”550 Mail from “ $&{client_addr} “ refused.mc file. Using Public Spam Blacklists with sendmail There are many publicly available lists of known open mail relay servers and spam generating mail servers on the Internet.ordb.html)”’.rfc-ignorant. the blacklists include not only the offending IP address but also the entire subnet or network block to which it belongs. You can configure sendmail to use its dnsbl feature to both query these lists and reject the mail if a match is found. others are managed by public companies. FEATURE(`dnsbl’.org/faq/\#why_rejected”’)dnl ☞ Spamcop: A spammer blacklist.spamhaus. `ipwhois.blackholes.easynet. Here are some sample entries you can add to your /etc/sendmail.nl’.

as in spamassassin-2.23Harrison_ch21. and restart spamassassin after booting: [root@bigboy tmp]# service spamassassin start [root@bigboy tmp]# service spamassassin stop [root@bigboy tmp]# service spamassassin restart Configuring procmail for spamassassin The /etc/procmailrc file is used by procmail to determine the procmail helper programs that should be used to filter mail. Downloading and Installing spamassassin Most Red Hat and Fedora Linux software products are available in the RPM format. it hands the message over to procmail. This file isn’t created by default. spamassassin doesn’t delete spam. stop.i386. it merely adds the word “spam” to the beginning of the subject line of suspected spam e-mails.rc. When searching for the RPMs.qxd 2/25/05 10:38 AM Page 359 Fighting Spam 359 spamassassin Once sendmail receives an e-mail message.rpm.) Starting spamassassin You can use the boot: chkconfig command to get spamassassin configured to start at [root@bigboy tmp]# chkconfig --level 35 spamassassin on To start. . You can make procmail temporarily hand over control to another program. remember that the filename usually starts with the software package name and is followed by a version number. which is the application that actually places the e-mail in user mailboxes on the mail server.60-2. Copy the template to the /etc directory: [root@bigboy tmp]# cp /etc/mail/spamassassin/spamassassin-spamc. such as a spam filter. You can then configure the e-mail filter rules in Outlook Express or any other mail client to either delete the suspect message or store it in a special Spam folder.rc /etc/procmailrc This file forces all mail arriving for your mail server’s users through Spamassasin. see Chapter 6. The most commonly used filter is spamassassin. spamassassin has a template you can use called /etc/mail/spamassassin/ spamassassin-spamc. (For help downloading.

spamassassin/user_prefs.english ok_languages en . The lower the # number the more sensitive it is.qxd 2/25/05 10:38 AM Page 360 360 Configuring Linux Mail Servers Chapter 21 Configuring spamassassin The spamassassin configuration local. 0=No) rewrite_subject 1 # Text to prepend to subject if rewrite_subject is used subject_tag *****SPAM***** # Encapsulate spam in an attachment (1=Yes. 0=No) report_safe 1 # Use terse version of the spam report (1=Yes.cf (see spamassassin(1) for details) # # How many hits before a message is considered spam. 0=No) use_terse_report 0 # Enable the Bayes system (1=Yes. 0=No) auto_learn 1 # Enable or disable network checks (1=Yes. required_hits 5.0 # Whether to change the subject of suspected spam (1=Yes. You can customize this fully meet your needs: file is named /etc/mail/spamassassin/ commented sample configuration file to ################################################################### # See ‘perldoc Mail::SpamAssassin::Conf’ for # details of what can be adjusted. ################################################################### # # These values can be overridden by editing # ~/. 0=No) use_bayes 1 # Enable Bayes auto-learning (1=Yes. # . 0=No) skip_rbl_checks 0 use_razor2 1 use_dcc 1 use_pyzor 1 # Mail using languages used in these country codes will not be marked # as being possibly spam in a foreign language.23Harrison_ch21.cf.

Update your .org).accept. .forward file and place an entry in /etc/smrsh. 4. The most important modules are: ☞ ☞ ☞ ☞ MailTools IO-Stringy MIME-tools Mail-Audit I have written a script called mail-filter. A few steps are required to make the script work: 1. This is fairly simple to do. Startup spamassassin The final steps are to configure start it.qxd 2/25/05 10:38 AM Page 361 Fighting Spam 361 # Mail using locales used in these country codes will not be marked # as being possibly spam in a foreign language. The sendmail program then looks for the filename in the directory /etc/smrsh and executes it.pl that effectively filters out spam e-mail for my home system. Place an executable version of the script in your home directory and modify the script’s $FILEPATH variable point to your home directory.cpan. By default. PERL doesn’t come with modules that are able to check e-mail headers and envelopes so you have to download them from CPAN (www. 2.reject. spamassassin to start on booting and then to [root@bigboy tmp]# chkconfig spamassassin on [root@bigboy tmp]# service spamassassin start Starting spamd: [ OK ] [root@bigboy tmp]# A Simple PERL Script to Help Stop Spam Blacklists won’t stop everything.23Harrison_ch21. which specifies the subjects and e-mail addresses to accept. ok_locales en Be sure to restart spamassassin for your changes to take effect. which specifies those to reject. Update file mail-filter. and file mail-filter. Install PERL and the PERL modules you downloaded from CPAN. 3. but you can limit the amount of unsolicited spam you receive by writing a small script to intercept your mail before it is written to your mailbox. because sendmail always checks the .forward file in your home directory for the name of this script.

It isn’t yet a part of the Fedora RPM set. You’ll also have to make your Linux box a POP mail server. Scripts. and Configurations.net. therefore. Installing Your POP Mail Server You need to install the imap RPM that contains the POP server software.net and featured in this section much more intuitive to use for the SOHO environment. As with all xinetd-controlled programs. you have to use the chkconfig command to make sure xinetd starts up on booting. It then denies everything else.org domain. For a simple script with instructions on how to install the PERL modules. [root@bigboy tmp]# chkconfig pop3 on To stop POP mail after booting. Remember that the filename is probably similar to imap-2002d-3. see Appendix II.” CONFIGURING YOUR POP MAIL SERVER Each user on your Linux box will get mail sent to their account’s mail folder. the chkconfig command also immediately activates application. then you have a few more steps. . “Codes. Microsoft Outlook or Outlook Express. If you want to retrieve the mail from your Linux box’s user account using a mail client such as Evolution. once again use chkconfig: [root@bigboy tmp]# chkconfig pop3 off Remember to restart the POP mail process every time you make a change to the configuration files to ensure the changes take effect on the running process.23Harrison_ch21. and you will probably have to download it from rpmfind. To configure POP mail to start at boot. Fedora Linux comes with its Cyrus IMAP/POP server RPM package. Starting Your POP Mail Server POP mail is started by xinetd.qxd 2/25/05 10:38 AM Page 362 362 Configuring Linux Mail Servers Chapter 21 Mail-filter first rejects all e-mail based on the reject file and then accepts all mail found in the accept file. but I have found the IMAP-2002 RPMs found on rpmfind. but sendmail just handles mail sent to your my-web-site.i386.rpm.

Encrypted POP runs on TCP port 995. for example—and create Linux accounts john1 and john2. Most POP clients support secure POP.org and john2@another-web-site.org). Windows configures it in the Advanced menu of the Outlook Express Account Configuration window. respectively.d/ipop3.org.23Harrison_ch21.qxd 2/25/05 10:38 AM Page 363 Configuring Your POP Mail Server 363 How to Configure Your Windows Mail Programs All your POP e-mail accounts are really only regular Linux user accounts in which sendmail has deposited mail.org) and John Brown (john@another-web-site. You can now configure your e-mail client such as Outlook Express to use your new POP/SMTP mail server quite easily. use /etc/xinetd. Troubleshooting POP Mail The very first troubleshooting step is to determine whether your POP server is accessible on the POP TCP port 110 or the secure POP port of 995.org entries to account john2. such as John Smith (john@my-web-site. You have two options for a solution: ☞ Make the user part of the e-mail address different—john1@my-website. To configure POP Mail. With this trick you’ll be able to handle many users belonging to multiple domains without many address overlap problems. then you may need to modify your virtusertable file. set your SMTP mail server to be the IP address/domain name of your Linux mail server. Lack of connectivity could be caused by a firewall with incorrect permit. Configuring Secure POP Mail If you need to access your e-mail from the mail server via the Internet or some other insecure location. both users will get sent to the Linux user account john by default. set your POP mail server to be the IP address of your Linux mail server.d/pop3s file instead of /etc/xinetd. The POP configuration in Outlook Express for each user should retrieve their mail via POP using john1 and john2. you may want to configure POP to work over an encrypted data channel. If the users insist on overlapping names. For example. or port .org to account john1 and point john@another-web-site. so firewall rules may need to be altered as well. NAT. ☞ Create the user accounts john1 and john2 and point virtusertable entries for john@my-web-site. For this. How to Handle Overlapping E-mail Addresses If you have user overlap. Use your Linux user username and password when prompted. Next.

16.org [172. “Troubleshooting with syslog.16.103] nmsgs=0/0 Aug 11 23:20:52 bigboy ipop3d[18694]: Logout user=labmanager host=17216-1-103. (Troubleshooting TCP with TELNET is covered in Chapter 4.) Linux status messages are logged to the file /var/log/messages.16.103 Aug 11 23:20:33 bigboy ipop3d[18693]: pop3s SSL service init from 172. your Web site won’t seem complete.103] nmsgs=0/0 Aug 11 23:20:40 bigboy ipop3d[18693]: Logout user=labmanager host=17216-1-103. Without it.my-web-site.1.1.” discusses many Linux-based tools that you can use to track the health of your Linux server.16.16.103] nmsgs=0 ndele=0 Aug 11 23:20:40 bigboy xinetd[18690]: EXIT: pop3s pid=18693 duration=7(sec) Aug 11 23:20:52 bigboy xinetd[18690]: START: pop3s pid=18694 from=172.my-web-site.1.103] nmsgs=0 ndele=0 Aug 11 23:20:52 bigboy xinetd[18690]: EXIT: pop3s pid=18694 duration=0(sec) CONCLUSION E-mail is an important part of any Web site.23Harrison_ch21. .”) Aug 11 23:20:33 bigboy xinetd[18690]: START: pop3s pid=18693 from=172.org [172. “Monitoring Server Performance. Check your configuration files if it fails to do so.1. Use it to make sure all your files are loaded when you start xinetd.103 Aug 11 23:20:52 bigboy ipop3d[18694]: pop3s SSL service init from 172.16. Test this from both inside your network and from the Internet. Chapter 22.1. (Linux logging is covered in Chapter 5. A fully functioning Web site is just the beginning. This example starts xinetd and makes a successful secure POP query from a remote POP client.103 Aug 11 23:20:40 bigboy ipop3d[18693]: Login user=labmanager host=17216-1-103. Failure could also be caused by the xinetd process being stopped or the configuration files being disabled.qxd 2/25/05 10:38 AM Page 364 364 Configuring Linux Mail Servers Chapter 21 forwarding rules to your server.103 Aug 11 23:20:52 bigboy ipop3d[18694]: Login user=labmanager host=17216-1-103.1. and you need to plan its configuration carefully to make it a seamless part of the Web experience of your visitors.16.org [172.my-web-site.1. It needs to be maintained to reduce the risk of failure and monitored to help detect potential problems.my-web-site.16.org [172.1.

For ease of use.qxd 2/25/05 10:14 AM Page 365 C H A P T E R 22 Monitoring Server Performance IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ SNMP MRTG Webalizer The top Command The vmstat Command The free Utility Conclusion You can monitor your system’s Web performance quite easily with graphical Linux tools. The tree starts from a root node. You’ll learn how to use several in this chapter. That’s a lot of new terms to digest in two sentences. including MRTG. routers and firewalls keep their operational statistics in object identifiers (OIDs) that you can remotely retrieve via the Simple Network Management Protocol (SNMP). which then descends through branches and leaves that each add their own reference 365 . and Webalizer. which tracks Web site hits. which is based on SNMP and monitors raw network traffic. SNMP Most servers. OIDs and MIBs OIDs are arranged in a structure of management information (SMI) tree defined by the SNMP standard. equipment vendors provide Management Information Base (MIB) files for their devices that define the functions of the OIDs they contain.24Harrison_ch22. so take a moment to look more closely.

internet.qxd 2/25/05 10:14 AM Page 366 366 Monitoring Server Performance Chapter 22 value to the path separated by a period.1 shows an OID structure in which the path to the enterprises OID branch passes through the org.6. The OID path for enterprises is.3. dod.4. therefore.1. 1. root node ccitt (0) iso (1) joint (2) org (3) dod (6) internet (1) directory (1) mgmt (2) experimental (3) private (4) enterprises (1) Figure 22.24Harrison_ch22. and private branches first. . Figure 22.1.1 The SNMP OID structure.

1 shows how some commonly used OIDs map to their MIB definitions. and private branches.6 1.2 1.1. you can reference the OID 1.109.5. which is named dod.1.1.109.6.1.3.1.9. the internet MIB under dod defines the function of the topmost OIDs in the directory.3. The Linux snmpget command outputs the value of a single leaf.3. all you need to know is that the command output frequently doesn’t list the entire OID.3. Equipment manufacturers are usually assigned their own dedicated OID branch under the enterprises MIB. For example. and they must also provide information in universally accepted OIDs for ease of manageability.3.3 1.1 1.4.1 1.3.1 OIDs and Their Equivalent MIBs MIB org dod internet directory mgmt experimental private enterprises You can refer to an OID by substituting the values in a branch with one of these more readable MIB aliases. Each branch is equivalent to a subdirectory.4.9.3.1.6.3 1. enabling you to click on an OID and see its value.4 1.4.1. for now. and the very last value at the tip (the leaf) correlates to a file containing data.qxd 2/25/05 10:14 AM Page 367 SNMP 367 Management Information Bases (MIBs) are text definitions of each of the OID branches. and the snmpwalk command provides the values of all leaves under a branch. experimental. For example. I’ll discuss these commands later.1.3. just the MIB file in which it was found and the alias within the MIB. Remember.1.3. For example: SNMPv2-MIB::sysUpTime.0 Here the OID value was found in the SNMPv2-MIB file and occupies position zero in the sysUpTime alias.1.1.6.1. the SMI org MIB defines all the topmost OIDs found at the next layer. mgmt.1.6.1). only the OID value at the very tip of a branch.1 OID 1. Table 22. type.6.1.6.9.1. Table 22. For example.9.6.24Harrison_ch22.6.5 as enterprises. the —leaf— actually has a readable value. This MIB information is very useful for SNMP management programs. NIC .1. Think of OIDs like the directory structure on a hard disk.1 by substituting the branch name (enterprises) for its OID numbers (1. and description.

Remember their differences. There are a number of types of community strings. often referred to as v2c. ☞ SNMP Version 3: This version provides greater security and remote configuration capabilities than its predecessors.qxd 2/25/05 10:14 AM Page 368 368 Monitoring Server Performance Chapter 22 interface data throughput values must always be placed in a predefined location in the general tree. expanded the number of supported error codes. The Read/Write or “set” community string is for not only viewing statistics and system parameters but also for updating the parameters. SNMP Versions There are currently three versions of SNMP: ☞ SNMP Version 1: The first version of SNMP to be implemented. but a memory use value on a customized processor card may be defined in a MIB under the manufacturer’s own OID branch. Support for encrypted SNMP data transfer and transfer error detection is also provided. Security was limited to community strings and access controls based on the IP address of the querying server. Views of OIDs in a MIB can also be limited on a per-user basis. In many cases the Read Only community string or password is set to the word “public”. SNMP v2c is backward compatible with version 1. Access isn’t limited to a single community string for read-only and read/write access. ☞ SNMP Version 2: The second version of SNMP. SNMP Community Strings As a security measure. you should change it from this easy-to-guess value whenever possible. Data communication wasn’t encrypted. because you will need to specify the version number when doing SNMP queries. increased the size of counters used to track data. as usernames and passwords have been introduced. and had the ability to do bulk queries that more efficiently loaded response packets with data. you need to know the SNMP password. . to query OIDs. version 1 was designed to be a protocol to provide device statistics and error reporting without consuming a lot of system resources.24Harrison_ch22. or community string. the most commonly used ones are the Read Only or “get” community string that only provides access for viewing statistics and system parameters.

.rpm. [root@bigboy tmp]# snmpwalk -v 1 -c craz33guy localhost interface .. When searching for the file..2 = STRING: 0:9:5b:2f:9e:d5 [root@bigboy tmp]# . IF-MIB::ifDescr.1-2... and 3. you can see that the second interface seems to have the name eth0 and the MAC address 0:9:5b:2f:9e:d5.. you need to have the SNMP utility tools package net-snmp-utils installed. [root@bigboy tmp]# Upon inspecting the output of the snmpwalk command. 2.1.2 = STRING: 0:9:5b:2f:9e:d5 IF-MIB::ifPhysAddress. Both commands require you to specify the community string with a -c operator.1 = STRING: lo IF-MIB::ifDescr.24Harrison_ch22. -v 2c. IF-MIB::ifPhysAddress.2 IF-MIB::ifPhysAddress. but it does require a number of detailed steps.1 = STRING: IF-MIB::ifPhysAddress. Installing SNMP Utilities on a Linux Server If you intend to use your Linux box to query your network devices. This example gets all the values in the interface MIB of the local server using SNMP version 1 and the community string of craz33guy. The first argument is the name or IP address of the target device and all other arguments list the MIBs to be queried. ..”) SNMP Utilities Command Syntax The SNMP utility tools package installs a number of new commands on your system for doing SNMP queries. You can now retrieve the individual MAC address using the snmpget command: [root@bigboy tmp]# snmpget -v 1 -c const1payted localhost ifPhysAddress.. other servers or even itself using MRTG or any other tool.. They also require you to specify the version of the SNMP query to be used with a -v 1...3 = STRING: eth1 .2 = STRING: eth0 IF-MIB::ifDescr. (If you need an installation refresher.. “Installing RPM Software. . see Chapter 6.. remember that the SNMP utility tools RPM’s filename usually starts with net-snmp-utils. which is followed by a version number.3 = STRING: 0:b0:d0:46:32:71 . respectively. or -v 3 operator for versions 1. most notably snmpget for individual OIDs and snmpwalk for obtaining the contents of an entire MIB.i386. as in net-snmp-utils5.qxd 2/25/05 10:14 AM Page 369 SNMP 369 Doing SNMP Queries Configuring SNMP on a server isn’t hard.

.qxd 2/25/05 10:14 AM Page 370 370 Monitoring Server Performance Chapter 22 You can confirm this information using the ifconfig command for interface eth0.0 = STRING: Linux bigboy 2.. installs the net-snmp package as its SNMP server product.255.240 .119. Save the old configuration file: [root@bigboy tmp]# cd /etc/snmp/ [root@bigboy snmp]# mv snmpd. the very first line shows a matching MAC address: [root@bigboy tmp]# ifconfig -a eth0 eth0 Link encap:Ethernet HWaddr 00:09:5B:2F:9E:D5 inet addr:216.255.conf. This package uses a configuration file named /etc/snmp/snmpd. The version of the configuration file that comes with net-snmp is quite complicated. I suggest archiving it and using a much simpler version with only a single line containing the keyword rocommunity followed by the community string. Configuring SNMP on a Linux Server By default Fedora.4.18-14 #1 Wed Sep 4 11:57:57 EDT 2002 i586 .10. . Enter this line in the new configuration file to set the Read Only community string to craz33guy: rocommunity craz33guy 3.10.conf in which the community strings and other parameters may be set.24Harrison_ch22.conf 2. Configure Linux to start SNMP services on each reboot with the chkconfig command: [root@bigboy root]# chkconfig snmpd on [root@bigboy root]# 4.conf snmpd..244 Bcast:216.. Start SNMP to load the current configuration file: [root@bigboy root]# service snmpd start Starting snmpd: [ OK ] [root@bigboy root]# 5.255 Mask:255. [root@bigboy tmp]# You’ll now see how you can configure SNMP on your Linux server to achieve these results.119. Here is an example: 1.old [root@bigboy snmp]# vi snmpd. Test whether SNMP can read the snmpwalk command: system and interface MIBs using the [root@bigboy snmp]# snmpwalk -v 1 -c craz33guy localhost system SNMPv2-MIB::sysDescr.

1 interface Different SNMP Versions There are currently three versions of SNMP.0 = STRING: bigboy . . [root@bigboy snmp]# snmpwalk -v 1 -c chir1qui 192.1. you were polling localhost.10 SNMPv2-MIB::sysUpTime.2 = INTEGER: 2 IF-MIB::ifIndex. to create online graphs of your traffic flows.0 = INTEGER: 3 IF-MIB::ifIndex.1..2 = STRING: wlan0 IF-MIB::ifDescr. Basic SNMP Security The most commonly supported versions of SNMP don’t encrypt your community string password so you shouldn’t do queries over insecure networks. Always make sure you are using the correct one.. Here is an example of a query of a device with an IP address of 192.conf file. .24Harrison_ch22.0 = STRING: root@localhost SNMPv2-MIB::sysName.1..0 = OID: NET-SNMP-MIB::netSnmpAgentOIDs.0 = Timeticks: (425) 0:00:04.. such as the Internet. 2 and 3.25 SNMPv2-MIB::sysContact.1 = STRING: lo IF-MIB::ifDescr... . The Linux snmpwalk and snmpget commands have -v 1.. You can also configure your server to use the TCP wrappers feature outlined in Appendix I. [root@bigboy snmp]# snmpwalk -v 1 -c craz33guy localhost interface IF-MIB::ifNumber.3 = INTEGER: 3 IF-MIB::ifDescr. [root@bigboy snmp]# Now that you know SNMP is working correctly on your Linux server. versions 1.. .168... -v 2c.qxd 2/25/05 10:14 AM Page 371 SNMP 371 SNMPv2-MIB::sysObjectID. All you need is the IP address and SNMP Read Only string and you’ll be able to get similar results. ..168.1 = INTEGER: 1 IF-MIB::ifIndex. and -v 3 switches for specifying the SNMP version to be used for queries. You should also make sure that you use all reasonable security measures to allow queries only from trusted IP addresses either via a firewall or the SNMP security features available in the snmp. you can configure SNMP statistics gathering software. such as MRTG.3 = STRING: eth0 . SNMP on Other Devices In the example.. You can poll any SNMP-aware network device that has SNMP enabled.

name ## ============== view all-mibs incl/excl ========= included MIB. .1. In case you need it. You will have to comment out the more restrictive statements at the top of the file for it to work correctly. and 192.0/24 192.24Harrison_ch22.level prefix read ========== ======= ========= ========= ====== ==== write ===== MyROGroup “” v1 noauth 0 all-mibs none In the example: ☞ Only three networks (localhost.name ================= MyROGroup MyROGroup MyROGroup sec.name context sec.16. ☞ Each network is matched to a group called MyROGroup using SNMP version 1. a section with very restrictive access sits at the top of the file and is immediately followed by a less restrictive section. ☞ An access statement ensures that only the defined networks have read only access to all the MIBs.name ======== local network_1 network_2 source ====== localhost 172.1. 172. The example that follows is a modification of the less restrictive section. ## ## com2sec com2sec com2sec sec.” to limit access to specific servers without the need of a firewall.name ======== local network_1 network_2 ## MIB.conf file has two security sections.0/24.subtree =========== . the snmpd.conf file can support limiting MIB access to trusted hosts and networks.168.1 mask ==== 80 ## ## notif ## ===== access none MIB group.view.1.16. The snmpd.16.0/24 community ========= craz33guy craz33guy craz33guy ## ## group group group Access.model sec.0/24) are allowed to access the server with the craz33guy community string.model ========= v1 v1 v1 sec.1.qxd 2/25/05 10:14 AM Page 372 372 Monitoring Server Performance Chapter 22 “Miscellaneous Linux Topics. and all the MIBs on the server are defined by the view named all-mibs.group.

2 shows a sample MRTG graph. MRTG MRTG (Multi-Router Traffic Grapher) is a public domain package for producing graphs of various router statistics via a Web page. ☞ Any TCP wrappers configuration on your SNMP target machine allows SNMP queries from your SNMP client.org) and also on your distribution CDs. then verify that: ☞ You restarted your snmp. You can easily create graphs of traffic flow statistics through your home network’s firewall/router or even your Linux box’s NIC cards using MRTG. A simple ping or traceroute test should be sufficient.allow) and (/etc/hosts. Figure 22. Simple SNMP Troubleshooting If your SNMP queries fail. ☞ The snmpd daemon is running on the SNMP client. ☞ You are using the correct community string. ☞ Firewalls aren’t preventing SNMP queries from the SNMP client to the SNMP target. ☞ You are querying using the correct SNMP version. Remember. the snmpd.deny) should be blank.conf file is only read by the snmpd daemon when it starts up. ☞ Network routing between the client and target devices is correct. These precautions are probably unnecessary in a home environment where access is generally limited to devices on the home network by an NAT firewall. ☞ Your SNMP security policy allows the query from your network. The product is available from the MRTG Web site (www. ☞ Your /var/log/messages file does not contain errors that may have occurred while starting snmpd.mrtg.24Harrison_ch22.conf file so the configuration settings become active. Troubleshooting to get functioning SNMP queries is important as many other supporting applications such as MRTG—which I’ll discuss next—rely on them in order to work correctly.qxd 2/25/05 10:14 AM Page 373 MRTG 373 ☞ Modification of the MIBs via SNMP is denied because the word “none” is in the write section of the access statement. . Generally speaking in a home environment protected by NAT your TCP wrappers files (/etc/hosts.

“The Apache Web Server. you can refer to Chapter 20. MRTG places its HTML files in /var/www/mrtg. By default Apache expects the HTML files for your Web site to be located in /var/www/html.2 A typical MRTG Web page. The easiest way to tell if Apache is installed is to run the rpm -q httpd command. Most Red Hat and Fedora Linux software products are available in the RPM format. remember that the MRTG RPM’s filename usually starts with mrtg and a version number.rpm.24Harrison_ch22. When searching for the file.qxd 2/25/05 10:14 AM Page 374 374 Monitoring Server Performance Chapter 22 Figure 22. as in mrtg-2. In addition to MRTG. Red Hat Linux usually comes with the Apache Web server software preinstalled.” for installation details.i386.5-3. If you don’t get a positive response.10. MRTG Download and Installation You need to install MRTG before proceeding. you need to install the SNMP utility tools as explained earlier and you need to have a Web server package installed for MRTG to work. .

.. Methods for specifying other OIDs.cfg \ --global “workdir: /var/www/mrtg” -ifref=ip \ --global ‘options[_]: growright..bits’ \ craz33guy@localhost --base: Get Device Info on craz33guy@localhost: --base: Vendor Id: --base: Populating confcache --snpo: confcache craz33guy@localhost: Descr lo --> 1 --snpo: confcache craz33guy@localhost: Descr wlan0 --> 2 . . . .. the errors prevent the graphs from being created correctly. “Advanced MRTG for Linux. Use MRTG’s cfgmaker command to create a configuration file named mrtg. MRTG maps the inbound and outbound data throughput rates on the device it is polling. this causes errors when Fedora Core 1 polls another Fedora Core 1 server.” For now. it creates a directory called /etc/mrtg in which all future configuration files are stored. --base: Walking ifAdminStatus --base: Walking ifOperStatus --base: Writing /etc/mrtg/mrtg. [root@bigboy tmp]# cfgmaker --output=/etc/mrtg/mrtg. You can avoid this problem by removing the option to get graph titles using the much less recognizable MAC address of the interfaces.. are discussed in Chapter 23.24Harrison_ch22.cfg [root@bigboy tmp]# Tip Using the -ifref=ip cfgmaker command option instructs MRTG to use the interface’s IP address as the title for each graph. To create a replacement default /etc/mrtg/mrtg. Place all data files in the directory /var/www/mrtg:.qxd 2/25/05 10:14 AM Page 375 MRTG 375 Configuring MRTG By default. When the MRTG RPM is installed. follow these steps: 1. I’ll stick with the default configuration.. such as CPU and memory usage.cfg configuration file for the server.cfg for the server (Bigboy) using a Read Only community string of craz33guy. Unfortunately.

cfg file as a guide.100.168. You can specify MRTG’s SNMP query version with the --snmp-options cfgmaker option.1.168.qxd 2/25/05 10:14 AM Page 376 376 Monitoring Server Performance Chapter 22 Tip As explained in the SNMP section. and the file that governs this is /etc/cron.24Harrison_ch22. 3. which doesn’t pass any network traffic at all.168. A certain candidate would be the virtual loopback interface Lo: (with the IP address of 127. Edit /etc/mrtg/mrtg.old updating log file [root@bigboy tmp]# env LANG=C /usr/bin/mrtg /etc/mrtg/mrtg.1).168.100 Rateup WARNING: /usr/bin/rateup The backup log file for localhost_192.1. Run MRTG using /etc/mrtg/mrtg.3. The --snmp-options option’s five colons before the 2 are important: [root@bigboy tmp]# cfgmaker -output=/etc/mrtg/192.168.1.html \ /etc/mrtg/mrtg.168.cfg as your argument three times. and naturally.d/mrtg. Use MRTG’s indexmaker command to create a Web index page using your new mrtg. replacing all occurrences of /usr/bin/mrtg with env LANG[eq]C /usr/bin/mrtg. If your query doesn’t work. .old updating log file [root@bigboy tmp]# env LANG=C /usr/bin/mrtg /etc/mrtg/mrtg.1.1. so the format of the command would be: [root@bigboy tmp]# indexmaker --output=/var/www/mrtg/index. MRTG is run every five minutes by default. Here is an example of cfgmaker using an SNMP version 2 query of a router with an IP address of 192.cfg 5.cfg \ -ifref=ip --global “workdir: /var/www/mrtg” \ --snmp-options=:::::2 craz33guy@192.100.1.0.3 2. edit this file.cfg Rateup WARNING: /usr/bin/rateup could not read the primary log file for localhost_192.log to localhost_192. there are different versions of SNMP.cfg Rateup WARNING: /usr/bin/rateup Can’t remove localhost_192. You’ll get an error the two times as MRTG tries to move old data files.3.0.100. The MRTG Web GUI expects to find the index file in the default MRTG Web directory of /var/www/mrtg/.cfg [root@bigboy tmp]# 4.168.100 was invalid as well Rateup WARNING: /usr/bin/rateup Can’t remove localhost_192.old updating log file Rateup WARNING: /usr/bin/rateup Can’t rename localhost_192. For MRTG to work correctly.1.168.168. the first time it is run.100. The explanation for changing the language character set for MRTG is given in the “Troubleshooting MRTG” section.1. MRTG has no data files to move.cfg. [root@bigboy tmp]# env LANG=C /usr/bin/mrtg /etc/mrtg/mrtg. check to make sure you are using the required version and then check other SNMP configuration parameters on the target device.1. and remove the sections related to interfaces you don’t need to monitor.

and then restart the cron daemon to re-read the configuration: [root@bigboy tmp]# service crond restart You could also create a script with the /usr/bin/mrtg/etc/mrtg/ entries in it and make cron run it every five minutes.cfg 0-59/5 * * * * root env LANG=C /usr/bin/mrtg /etc/mrtg/device2. Other versions of Linux keep their MRTG cron entries inside the /etc/crontab file. Edit this file using the same syntax as the Fedora /etc/cron. choose a filename that is not mrtg.cfg 3. Create a new configuration file using the steps from the previous section.cfg /etc/mrtg/device2.conf that includes all the necessary Apache commands for MRTG to work.d/mrtg.1.d/mrtg. device. With Fedora Core.cfg 4.cfg Configuring Apache to Work with MRTG MRTG is useful because it can provide a graphical representation of your server’s performance statistics via a Web browser.cfg /etc/mrtg/device1. This will be covered later.cfg file every five minutes to poll your network devices.html \ /etc/mrtg/mrtg. or you can allow universal access by commenting out that line along with the Deny from line. and include all of your ration files. Add a new MRTG line in /etc/cron. but first I’ll show you how to poll multiple devices. This way you can just edit the script each time you add a device without having to restart cron.cfg 0-59/5 * * * * root env LANG=C /usr/bin/mrtg /etc/mrtg/device1. MRTG creates an add-on configuration file named /etc/httpd/conf.168.24Harrison_ch22. /etc/mrtg configu- [root@bigboy tmp]# indexmaker --output=/var/www/mrtg/index. each with a separate configuration file. You can configure MRTG to poll multiple devices. This file tells the cron daemon to run MRTG using the /etc/mrtg/mrtg. This example adds access from the 192.cfg. Run the indexmaker command. You can add your home network to the file by inserting the network on the Allow from line.d/mrtg for each new configuration file you create: 0-59/5 * * * * root env LANG=C /usr/bin/mrtg /etc/mrtg/mrtg. because by default MRTG accepts Web requests from the Linux console only. Some configuration may need to be done. to regenerate your Web index page.qxd 2/25/05 10:14 AM Page 377 MRTG 377 This isn’t all—you need to view the graphs too. 2. Here’s how: 1.0 network: .d/mrtg file. Getting MRTG to Poll Multiple Devices The Fedora Core MRTG installation process creates a cron file named /etc/cron.

d/mrtg.conf” Basic Security If you are accessing MRTG graphs from the Internet. .24Harrison_ch22. you may want to add password protection to the directory by using a . CPU. then you’ll have to comment out the Deny statement and allow from all IP addresses: <Location /mrtg> Order deny. With Fedora Core 1.qxd 2/25/05 10:14 AM Page 378 378 Monitoring Server Performance Chapter 22 <Location /mrtg> Order deny. Apache automatically reads the add-on files in the /etc/httpd/conf.org. weekly. Chapter 23 has detailed examples and explanations of how to monitor Linux disk. and yearly statistics for your interfaces.conf to find it. you have to specifically configure the Apache configuration file /etc/httpd/conf/httpd. monthly. The MRTG Web site. How to View the MRTG Graphs in Your Web Browser You can now access your MRTG graphs by pointing your browser to the URL: http://server-ip-address/mrtg/ Using MRTG to Monitor Other Subsystems MRTG will generate HTML pages with daily. Note With newer versions of Fedora. also has links to other sites that show you how to monitor many other subsystems on a variety of devices and operating systems.mrtg.1.0/24 </Location> If you want to access MRTG from the Internet.allow Allow from all </Location> Remember to restart Apache after you make these modifications for the changes to take effect.168.htaccess file as described in Chapter 20. By default. memory.d/ directory. www. You can do this yourself by inserting this line at the very bottom of the main Apache configuration file before restarting Apache for the change to take effect: include “/etc/httpd/conf. and Web connection data.allow Deny from all Allow from localhost 192. MRTG provides only network interface statistics.

and the current Fedora default of en_US. which is the much harder to recognize interface MAC address.qxd 2/25/05 10:14 AM Page 379 MRTG 379 Troubleshooting MRTG There are many simple steps you can use to troubleshoot MRTG. ------------------------------------------------------------------[root@bigboy tmp]# Your system’s character set is defined in /etc/sysconfig/i18n.. This is not necessarily a good idea. If it isn’t set. other things may fail to work.. where this is not the case: env LANG=C /usr/bin/mrtg . Fedora Core 1 MRTG Errors with net-snmp A bug appears in the MRTG implementation for some Fedora Core 1 MRTG versions when polling another Fedora Core 1 server. Make sure you follow the SNMP troubleshooting steps if you have any difficulties.cfg ------------------------------------------------------------------ERROR: Mrtg will most likely not work propperly when the environment variable LANG is set to UTF-8. then you’ll get errors such as this every time you run MRTG from the command line or as part of a cron job: root@bigboy tmp]# mrtg /etc/mrtg/mrtg.. Upgrading to the most current Core 1 version of MRTG will fix the problem: . because it makes MRTG provide graphs sorted by the IP addresses of the interfaces instead of the default. this statement is very convenient. Take a look at some of the most common ones. Basic Steps MRTG won’t work if SNMP queries don’t work. Setting the Correct Character Set MRTG usually works only if your system uses an ASCII-based (Western European) character set. but en_US will after a system reboot.24Harrison_ch22. A better solution is to always run MRTG using this command instead of using just plain /usr/bin/mrtg: env LANG=C /usr/bin/mrtg This will modify the character set used by MRTG alone and shouldn’t affect anything else.UTF-8 won’t work. When using a -ifref=ip statement with the cfgmaker command. every line in the configuration file that is generated becomes commented out. especially if the native language Linux uses on your system is not ASCII based. When it works. Please run mrtg in an envir.

html /etc/mrtg/mrtg.168.100’ ### ### The following interface is commented out because: ### * has a speed of which makes no sense ### * got ‘Received SNMP response with error code ### error status: noSuchName ### index 1 (OID: 1.1.168.qxd 2/25/05 10:14 AM Page 380 380 Monitoring Server Performance Chapter 22 ### Interface 6 >> Descr: ‘’ | Name: ‘’ | Ip: ‘192.100:craz33guy@localhost: # SetEnv[localhost_192.100]: MRTG_INT_IP=”192. [root@bigboy tmp]# WEBALIZER Webalizer is a Web server log file analysis tool that comes installed by default on Red Hat/Fedora Linux.2.1.1.100]: Traffic Analysis for 192.1.0.168.100” MRTG_INT_DES # MaxBytes[localhost_192.100]: 0 # Title[localhost_192.1.1]. Each night.html file and gives errors: [root@bigboy tmp]# indexmaker --output=/var/www/mrtg/stats/index.161) ### community: “craz33guy” ### request ID: 824482716 ### PDU bufsize: 8000 bytes ### timeout: 2s ### retries: 5 # # Target[localhost_192.168.168.168.3.0 bits/s</TD></TR> # </TABLE> As all the lines in the configuration file are commented out with a # character.168.2.100]: /192.24Harrison_ch22. Webalizer reads your Apache log files and creates a set of Web pages that enable you to view Web surfer statistics for your site.100]: <H1>Traffic Analysis for 192.168.10.168.6) ### SNMPv1_Session (remote host: “localhost” [127.1.1. The information provided includes a list of your Web site’s most popular pages sorted by hits along with traffic graphs showing the times of day when your site is most popular.1.168.0. indexmaker fails to create an index.2.1.100 # <TABLE> # <TR><TD>System:</TD> <TD>bigboy in Unknown</TD></TR> # <TR><TD>Maintainer:</TD> <TD>root@localhost</TD></TR> # <TR><TD>Description:</TD><TD> </TD></TR> # <TR><TD>ifType:</TD> <TD> ()</TD></TR> # <TR><TD>ifName:</TD> <TD></TD></TR> # <TR><TD>Max Speed:</TD> <TD>0.1.1.cfg Use of uninitialized value in hash element at /usr/bin/indexmaker line 307.100 # PageTop[localhost_192.6.1. .1.

d/webalthat includes all the necessary Apache commands for Webalizer to work.0% hi. 0.24Harrison_ch22. be aware that running in quiet mode could hide deeper problems that could occur in the future.32 ksoftirqd/0 3 root 5 -10 0 0 0 S 0. Webalizer places its index page in the directory /var/www/html/usage and allows you to view your data by visiting the URL http://server-ip-address/usage.0 0:00.1% wa.0% ni.conf.0 0. however. 80k used. 0. By default. 2 users.3% us.00 Tasks: 72 total.02 khelper [root@bigboy tmp]# .1% id. load average: 0. you have to edit it to allow access to the Webalizer pages from locations other than the Linux console.0 0:00. 356256k PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 1 root 16 0 2268 516 1420 S 0. After adjustments.qxd 2/25/05 10:14 AM Page 381 The top Command 381 How to View Your Webalizer Statistics izer. You also have to restart Apache to make the changes take effect. As in the case of the MRTG add-on file mentioned above. 0.0 0. 0 zombie Cpu(s): 4.00.1 0:03. Mem: 1027784k total. 95.85 init 2 root 34 19 0 0 0 S 0. This can be adjusted with the OutputDir directive in the file. 1 running. but you may want to adjust the directory in which Webalizer places your graph statistics.5% sy. Webalizer functions with few annoyances. 0. 159300k free.0 0. 71 sleeping. 0 stopped.00 events/0 4 root 7 -10 0 0 0 S 0.0 0. The Webalizer Configuration File Webalizer stores its configuration in the file /etc/webalizer. 276272k Swap: 4192924k total.conf Fedora creates an add-on configuration file named /etc/httpd/conf. THE TOP COMMAND You can monitor the amount of memory and CPU resources your system is using with the top command: [root@bigboy tmp]# top top .0 0:00.20:00:54 up 61 days. 0. 4192844k free. 868484k used. The default settings should be sufficient for your Web server. 23:35.

196MB used from a total of 130.780) and only 80MB of the swap partition is being used. memory is still 15% free (158. The amount of free memory may appear low. [root@bigboy tmp]# vmstat procs ------memory----.---io-. but in this case.-system. Excess swapping of programs and data between disk and memory can cause your system to slow down significantly and memory usage should be monitored to allow you to plan ways to either increase RAM or tune the way your system operates. Excessive swapping can cause your system to slow down dramatically. the server doesn’t seem to be swapping idle processes from memory to the swap disk partition as it isn’t being used at all. In the example. If your system seems slow but the CPU and memory usage is low. The output is easier to understand than vmstat’s. and network congestion due to excessive traffic. such as poor duplex negotiation. bad cables. the simplest ways to avoid this is to add more RAM or reduce the number of processes or users that are active on your system.--cpu-r b swpd free buff cache si so bi bo sy id wa 0 0 80 158196 276272 357236 0 0 1 0 1 95 0 [root@bigboy tmp]# in 0 cs us 4 4 As your memory fills up.—-swap. your system temporarily stores programs and data on your hard disk’s swap partition.1% idle and 15% of memory (159300k) is free. THE FREE UTILITY The free utility can determine the amount of free RAM on your system.qxd 2/25/05 10:14 AM Page 382 382 Monitoring Server Performance Chapter 22 Here the CPU usage is running 95. which provides a summary of what top produces. then start looking at networking problems. THE VMSTAT COMMAND You can also determine memory and swap usage with the vmstat command.24Harrison_ch22. Here’s a sample: [root@bigboy tmp]# vmstat procs —————-memory————— —-swap— ——-io—— r b swpd free buff cache si 0 0 80 158196 276272 357236 0 [root@bigboy tmp]# so 0 bi 1 bo 0 .

and vmstat. but also to graph many of the statistics listed in top. If you cannot achieve this. upgrading the capacity of your RAM. free. offloading some applications to servers with less load. which should allow it to handle moderate spikes in usage caused by running memory-intensive cron batch jobs or tape backups. of course.24Harrison_ch22. Chapter 23 shows you how. because it can help you predict when things are going to go wrong or long-term trends in your Web traffic. consider running more efficient versions of programs. and. . CONCLUSION Server monitoring is always a good practice. MRTG can be expanded not only to monitor traffic on your server’s NIC cards.qxd 2/25/05 10:14 AM Page 383 Conclusion 383 You should generally try to make your system run with at least 20% free memory on average.

24Harrison_ch22.qxd 2/25/05 10:14 AM Page 384 .

You may also want to see graphs of CPU. “Monitoring Server Performance”). 385 . MIBs are data structures that are constantly updated via the SNMP daemon. All the chapter’s examples assume that the SNMP Read Only string is craz33guy and that the net-snmp-utils RPM package is installed (see Chapter 22.qxd 2/25/05 10:15 AM Page 385 C H A P T E R 23 Advanced MRTG for Linux IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Locating and Viewing the Contents of Linux MIBs Testing Your MIB Value Differences in MIB and MRTG Terminology The CPU and the Memory Monitoring MIB The TCP/IP Monitoring MIB Manually Configuring Your MRTG File Implementing Advanced Server Monitoring Monitoring Non-Linux MIB Values Troubleshooting Conclusion In many cases using MRTG in a basic configuration to monitor the volume of network traffic to your server isn’t enough.25Harrison_ch23. LOCATING AND VIEWING THE CONTENTS OF LINUX MIBS Residing in memory. and memory usage. disk. This chapter explains how to find the values you want to monitor in the SNMP MIB files and then how to use this information to configure MRTG. The MIB configuration text files are located on your hard disk and loaded into memory each time SNMP restarts.

Don’t change them... tcpActiveOpens OBJECT-TYPE SYNTAX Counter ACCESS read-only STATUS mandatory DESCRIPTION “The number of times TCP connections have made a direct transition to the SYN-SENT state from the . MIBs are very complicated. RFC1213-MIB..txt file.0.txt: “The number of times TCP connections have made a .. because doing so could cause SNMP to fail. Each value tracked in a MIB is called an object and is often referred to by its object ID or OID.. TCP-MIB.txt: “The number of times TCP connections have made a .. [root@silent mibs]# You can use the vi editor to look at the MIBs.25Harrison_ch23. Integer values change instant by instant and are useful for tracking such statistics as the amount of memory currently being used. ..qxd 2/25/05 10:15 AM Page 386 386 Advanced MRTG for Linux Chapter 23 You can easily find your Fedora Linux MIBs by using the locate command and filtering the output to include only values with the word “snmp” in them. the MIBs are located in the /usr/share/snmp/mibs directory: [root@bigboy tmp]# locate mib | grep snmp /usr/share/doc/net-snmp-5.txt .. MIBs usually track two types of values. As you can see in this case. you can see that querying the tcpActiveOpens object returns the number of active open TCP connections to the server. .txt: “The limit on the total number of TCP connections RFC1213-MIB..6/README. This example searches for the MIBs that keep track of TCP connections and returns the RFC1213 and TCP MIBs as the result. [root@silent mibs]# grep -i tcp /usr/share/snmp/mibs/*. In this snippet of the RFC1213-MIB. The SYNTAX field shows that this is a counter value. such as the amount of packets passing through a NIC or amount of time the CPU has been busy since boot time..mib2c /usr/share/snmp/mibs /usr/share/snmp/mibs/DISMAN-SCHEDULE-MIB. [root@bigboy tmp]# As the MIB configurations are text files you can search for keywords in them using the grep command.txt \ | grep connections . but fortunately the key sections are commented. Counter values are used for items that continuously increase as time passes...

Follow up with the snmpwalk command shows that the value is located in ssCpuRawUser. snmpget are the ones you DIFFERENCES IN MIB AND MRTG TERMINOLOGY Always keep in mind that MRTG refers to MIB counter values as counter values. the ssCpuRawUser OID alias was found to be interesting. The snmpget is then successful in retrieving the counter32 type data with a current value of 396271. It refers to MIB integer and gauge values as gauge. giving an error stating that the MIB variable couldn’t be found. Understanding them will be important in understanding how to use MRTG to track MIB values. you can use the snmpwalk command to poll it. Failed object: UCD-SNMP-MIB::ssCpuRawUser [root@bigboy tmp]# [root@bigboy tmp]# snmpwalk -v1 -c craz33guy localhost ssCpuRawUser UCD-SNMP-MIB::ssCpuRawUser. TESTING YOUR MIB VALUE Once you identify an interesting MIB value for your Linux system. By default. CPU usage is typically tracked by MIBs as a counter value. MRTG considers all values to be counters. fortunately. Many times the text aliases in an MIB reference the OID branch only and not the data located in a leaf ending in an additional number.1. but the snmpget command fails to get a value. The snmpget command doesn’t work with branches. such as .” ::= { tcp 5 } You’ll explore the differences between SNMP and MRTG terminologies in more detail later.0 = Counter32: 396241 [root@bigboy tmp]# snmpget -v1 -c craz33guy localhost ssCpuRawUser.0 = Counter32: 396271 [root@bigboy tmp]# The MIB values that work successfully with should use with MRTG.25Harrison_ch23. it plots only how much the value has changed since the last polling cycle. . [root@bigboy tmp]# snmpget -v1 -c craz33guy localhost ssCpuRawUser Error in packet Reason: (noSuchName) There is no such variable name in this MIB. MRTG doesn’t plot counter values as a constantly increasing graph. In the example below. you can edit your MRTG configuration file to make it graph this information in a percentage use format (more on this later).0 UCD-SNMP-MIB::ssCpuRawUser.0 or .0 instead.qxd 2/25/05 10:15 AM Page 387 Differences in MIB and MRTG Terminology 387 CLOSED state.

the MIB object name. Table 23.qxd 2/25/05 10:15 AM Page 388 388 Advanced MRTG for Linux Chapter 23 The syntax type. and nice values can give a good approximation of total CPU usage.25Harrison_ch23. THE CPU AND THE MEMORY MONITORING MIB The UCD-SNMP-MIB MIB keeps track of a number of key performance MIB objects. and the description of what it does are the most important things you need to know when configuring MRTG. Adding the user.1. Subtracting this value from 100 can give a good approximation of total CPU usage. Table 23. The percentage of the time the CPU is running idle. Total CPU usage by applications running at a nondefault priority level.1 Important Objects in the UCD-SNMP-MIB MIB MIB Type Counter MRTG Type Counter Description Total CPU usage by applications run by nonprivileged users since the system booted. Total CPU usage by applications run by privileged system processes since the system booted.2 details the most important objects in TCP-MIB. UCD-SNMP-MIB Object Variable ssCpuRawUser ssCpuRawSystem Counter Counter ssCpuRawNice Counter Counter ssCpuRawIdle Counter Counter memAvailReal Integer Gauge THE TCP/IP MONITORING MIB The TCP-MIB MIB keeps track of data connection information and contains the very useful tcpActiveOpens and tcpCurrEstab objects. including the commonly used ones in Table 23. Available physical memory space on the host. . I’ll come back to these later. system.

The indexmaker program also uses this information to create the overview. the first one is the input OID. Measures the number of TCP connections in the established state. The mrtg program then uses these configuration files to determine the type of data to record in its data directory. The two OID values are separated by an & symbol. Parameter Formats MRTG configuration parameters are always followed by a graph name surrounded by square brackets and a colon. monthly. others refer to the Summary View.25Harrison_ch23.qxd 2/25/05 10:15 AM Page 389 Manually Configuring Your MRTG File 389 Table 23. because cfgmaker isn’t aware of any OIDs other than those related to a NIC. or Summary View Web page for the MIB OIDs you’re monitoring. simultaneously tracking two OIDs: the NIC’s input and output data statistics. . The format looks like this: Parameter[graph name]: value For ease of editing. and the second one is the output OID. and annual graphs. weekly. This Summary View page shows daily statistics only. You have to click on the Summary View graphs to get the Detailed View page behind it with the daily. If you want to monitor any other pairs of OIDs. Each graph can track two OIDS listed in the Target parameter. Total number of TCP segments with bad checksum errors. Some of the parameters in the configuration file refer to the Detailed View.2 Variable Important Objects in the TCP-MIB MIB MIB Type Counter Gauge Counter MRTG Type Counter Gauge Counter Description Measures the number of completed TCP connections. you have to manually create the configuration files. the parameters for a particular graph are usually grouped together. which is usually placed at the very top of the graph name list. TCP-MIB Object tcpActiveOpens tcpCurrEstab tcpInErrs MANUALLY CONFIGURING YOUR MRTG FILE The MRTG cfgmaker program creates configuration files for network interfaces only. The mrtg and indexmaker program can be fed individual OIDs from a customized configuration file and will function as expected if you edit the file correctly.

and legend2 is the expansion of legendO. Options Parameters Options parameters provide MRTG with graph formatting information.qxd 2/25/05 10:15 AM Page 390 390 Advanced MRTG for Linux Chapter 23 Legend Parameters On the Detailed View Web page. each graph has a legend that shows the max. MRTG defaults to growing from the left. This usually makes the graphs more intuitively easy to read. Likewise. and current values of the graph’s OID statistics.perminute options[graph2]: gauge. Here is an example of the legends of a default MRTG configuration: YLegend[graph1]: Legend1[graph1]: Legend2[graph1]: LegendI[graph1]: LegendO[graph1]: Bits per second Incoming Traffic in Bits per Second Outgoing Traffic in Bits per Second In Out You can prevent MRTG from printing the legend at the bottom of the graph by leaving the value of the legend blank like this: LegendI[graph1]: Later you’ll learn how to match the legends to the OIDs for a variety of situations. The growright option makes sure the data at the right of the screen is for the most current graph values.25Harrison_ch23. you can convert per second values to per minute graphs using the perminute option.bits . In the case of a default MRTG configuration this would be the data flow through the interface in bits or bytes per second. Parameter legend1 is the expansion of legendI. average. the value you are trying to compare. Here are some examples for two different graphs: options[graph1]: growright. The space available under each graph’s legend is tiny so MRTG also has legend1 and legend2 parameters that are placed at the very bottom of the page to provide more details. If the value you are monitoring is in bytes. You can use the legendI parameter for the description of the input graph (first graph OID) and the legendO for the output graph (second graph OID). The Ylegend is the legend for the Y axis. The gauge option alerts MRTG to the fact that the graphed values are of the gauge type. The nopercent option prevents MRTG from printing percentage style statistics in the legends at the bottom of the graph. then you can convert the output to bits using the bits option.nopercent.

the PageTop parameter tells the title for the Detailed View page. you may want to turn off MRTG’s auto scaling. MRTG always tries to compare two MIB OID values that are defined by the Target parameter. Title[graph1]: Interface eth0 PageTop[graph1]: <H1>Detailed Statistics For Interface eth0</H1> Scaling Parameters The MaxBytes parameter is the maximum amount of data MRTG will plot on a graph. so that the vertical peak is near the top of the graph image. and the server reaches a maximum of 60%. Anything more than this seems to disappear over the edge of the graph.0&mib-object-2. then the peak will be only 60% of the way up as the graph plots from 0% to 100%. When you are plotting a value that has a known maximum and you always want to have this value at the top of the vertical legend. MRTG will have a vertical plot of 0% to 60%. The PageTop string must start with <H1> and end with </H1>. weekly. You have to specify the two MIB OID objects—the SNMP password and the IP address of the device you are querying in this parameter—and separate them with an & character: Target[graph1]: mib-object-1. and daily views on the Detailed View page and gives them a maximum value of 100: Unscaled[graph1]: ymwd MaxBytes[graph1]: 100 Defining The MIB Target Parameters As stated before.25Harrison_ch23. This is so even if you set the MaxBytes parameter.qxd 2/25/05 10:15 AM Page 391 Manually Configuring Your MRTG File 391 If you place this parameter at the top with a label of [_] it gets applied to all the graphs defined in the file: options[_]: growright Title Parameters The title on the Summary Page is provided by the Title parameter. The example removes scaling from the yearly. If you are plotting percentage CPU usage. MRTG also tries to adjust its graphs so that the largest value plotted on the graph is always close to the top.0:<SNMP-password>@<IP-address> . When scaling is off. with scaling. and MaxBytes is set to 100. monthly.

Target[graph1]: mib-object-1. legend1 and legendI describe and legend2 and legendO describe mib-object-2.0:<SNMP-password>@<IPaddress> + mib-object-3. Table 23.0 for the other.0:<SNMP-password>@<IPaddress> Adding MIB Values Together for a Graph You can use the plus sign between the pairs of MIB object values to add them together.0 is tagged onto the end of the output.0&mib-object-1. in this case . The next example adds mib-object-1. Notice how the . The next example uses the SNMP command to return the user mode CPU utilization of a Linux server. you can just repeat the target MIB in the definition as in the next example.0.0.0:<SNMP-password>@<IPaddress> Plotting Only One MIB Value If you want to plot only one MIB value.0 for one graph and adds mib-object-2. mib-object-1. which plots only mib-object-1.0 and mib-object-4. Target[graph1]: mib-object-1.3.qxd 2/25/05 10:15 AM Page 392 392 Advanced MRTG for Linux Chapter 23 The numeric value.0 = Counter32: 926739 [root@silent mibs]# The MRTG legends map to the MIBs listed in the target as shown in Table 23.0:<SNMP-password>@<IP-address> .3 Legend Legend1 Legend2 LegendI LegendO Mapping MIBs to the Graph Legends Maps to Target MIB 1 2 1 2 So in the example below.0 and mib-object-3.0&mib-object-2.0&mib-object-2. at the end of the MIB is required.0.0&mib-object-4.25Harrison_ch23. The resulting MRTG graph actually superimposes the input and output graphs one on top of the other. Target[graph1]: mib-object-1. [root@silent mibs]# snmpwalk -v 1 -c craz33guy localhost ssCpuRawUser UCD-SNMP-MIB::ssCpuRawUser.

There must be white spaces before and after all these operators for MRTG to work correctly. Left and right parentheses are also valid. On smaller Web sites you may want to use the perminute option to make the graphs more meaningful: Target[graph1]: tcpPassiveOpens. multiply (*). If not. The tcpActiveOpens MIB object monitors new connections originating from the server. user mode processes. This example adds them all together in a single plot: Target[graph1]:ssCpuRawUser. Notice how the mandatory white spaces separate the mathematical operators from the next target element.0:<SNMP-password>@<IPAddress> ) * 100 / ( memTotalReal. Target[graph1]: ( memAvailReal.0:<SNMPpassword>@<IP-address> MaxBytes[graph1]: 1000000 Options[graph1]: perminute .0&ssCpuRawNice.0&memTotalReal. Sample Target: Memory Usage Here is an example for plotting the amount of free memory versus the total RAM installed in the server. The tcpPassiveOpens MIB object tracks newly created connections and is suited for this type of data transfer.0&memTotalReal. such as subtract (-).0& tcpPassiveOpens.gauge Sample Target: Newly Created Connections HTTP traffic caused by Web browsing usually consists of many very shortlived connections.gauge Next.growright. Sample Target: Total CPU Usage Linux CPU usage is occupied by system processes.0& memAvailReal. and divide (%).growright. Notice that this is a gauge type variable: Target[graph1]: memAvailReal. you’ll get oddly shaded graphs.0&ssCpuRawUser.0:<SNMP-password>@<IPaddress> + ssCpuRawSystem.qxd 2/25/05 10:15 AM Page 393 Manually Configuring Your MRTG File 393 You can use other mathematical operators.0:<SNMP-password>@<IPaddress> + ssCpuRawNice.25Harrison_ch23. and a few processes running in nice mode.0:<SNMPpassword>@<IP-Address> ) options[graph1]: nopercent. plot the percentage of available memory.0:<SNMP-password>@<IP-address> Be sure to place this command on a single line.0&ssCpuRawSystem.0:<SNMP-password>@<IPaddress> options[graph1]: nopercent.

2 = INTEGER: 73 [root@bigboy tmp]# Your MRTG target for these gauge MIB objects should look like this: Target[graph1]: dskPercent. create longer established connections while people download large files or stay logged into the server. Object (/home).1:<SNMP-password>@<IPaddress> options[graph1]: growright. Add two entries to your snmpd.conf dskPercent. Restart the SNMP daemon to reload the values: [root@bigboy tmp]# service snmpd restart 4.0:<SNMP-password>@<IPaddress> MaxBytes[graph1]: 1000000 Options[graph1]: gauge Sample Target: Disk Partition Usage In this example. Use the snmpwalk command to query the the dskPercent dskPercent.conf file: disk disk /home /var 3.1 refers to the first disk entry in snmpd. you’ll monitor the /var and /home disk partitions on the system: 1. and [root@bigboy tmp]# snmpwalk -v 1 -c craz33guy localhost \ dskPercent.25Harrison_ch23.gauge .1 UCD-SNMP-MIB::dskPercent.2 UCD-SNMP-MIB::dskPercent. MIB.qxd 2/25/05 10:15 AM Page 394 394 Advanced MRTG for Linux Chapter 23 Sample Target: Total TCP Established Connections Other protocols. The tcpCurrEstab MIB object measures the total number of connections in the established state and is a gauge value: Target[graph1]: tcpCurrEstab.0&tcpCurrEstab. First use the df -k command to get a list of the partitions in use: Used Available Use% Mounted 128199 19178 122864 8229 890092 171832 351302 76692 860188 471272 2832140 1261520 27% 21% 13% 2% 24% 73% / /boot /home /tmp /usr /var [root@bigboy tmp]# df -k Filesystem 1K-blocks on /dev/hda8 505605 /dev/hda1 101089 /dev/hda5 1035660 /dev/hda6 505605 /dev/hda3 3921436 /dev/hda2 1510060 [root@bigboy tmp]# 2. such as FTP and SSH.1 = INTEGER: 13 [root@bigboy tmp]# snmpwalk -v 1 -c craz33guy localhost \ dskPercent.1& dskPercent.2 refers to the second (/var).

25Harrison_ch23.cpu]:ssCpuRawUser. and then you can integrate it into the existing MRTG configuration.qxd 2/25/05 10:15 AM Page 395 Implementing Advanced Server Monitoring 395 Defining Global Variables You have to make sure MRTG knows where the MIBs you’re using are located. /usr/share/snmp/mibs/TCP-MIB. memory.txt.cfg # # Configuration file for non bandwidth server statistics # localhost for # # Define global options # LoadMIBs: /usr/share/snmp/mibs/UCD-SNMPMIB.System.cpu]: Server CPU Load PageTop[server.0&ssCpuRawUser. disk.txt workdir: /var/www/mrtg/ # # CPU Monitoring # (Scaled so that the sum of all three values doesn’t exceed 100) # Target[server.txt workdir: /var/www/mrtg/ IMPLEMENTING ADVANCED SERVER MONITORING You now can combine all you have learned to create a configuration file that monitors all these variables.0&ssCpuRawNice. You must also define where the HTML files will be located./usr/share/snmp/mibs/TCP-MIB.0:craz33guy@localhost + ssCpuRawNice. Specify their locations with the global LoadMIBs parameter. The default location MRTG uses may not be valid. A Complete Sample Configuration Here is a sample configuration file that is used to query server CPU.txt. and TCP connection information: # # File: /etc/mrtg/server-info. User and Nice Processes</H1> .cpu]: <H1>CPU Load .0:craz33guy@localhost + ssCpuRawSystem. the example specifies the default Fedora MRTG HTML directory: LoadMIBs: /usr/share/snmp/mibs/UCD-SNMP-MIB.0:craz33guy@localhost Title[server.0&ssCpuRawSystem.

T.memory]: Free LegendO[server.0&memAvailReal.qxd 2/25/05 10:15 AM Page 396 396 Advanced MRTG for Linux Chapter 23 MaxBytes[server.mempercent]: PageTop[server. not including swap.memory]: Total memory Options[server.mempercent]: Free Legend1[server.0&memTotalReal.cpu]: Options[server.newconns]: <H1>New TCP Connections</H1> MaxBytes[server.0:craz33guy@localhost Title[server.cpu]: % YLegend[server.memory]: memAvailReal.M.cpu]: Current CPU percentage load LegendI[server.newconns]: c/s .mempercent]: Percent LegendI[server.mempercent]: Percentage Free Memory # # New TCP Connection Monitoring (per minute) # Target[server.memory]: B YLegend[server.0&tcpActiveOpens.memory]: Free memory.mempercent]: 100 YLegend[server.mempercent]: Percentage Free Memory Legend2[server.nopercent Unscaled[server.memory]: k.memory]: Total Legend1[server.X # # Memory Monitoring (Percentage usage) # Title[server.nopercent kMG[server.transparent.cpu]: ymwd # # Memory Monitoring (Total Versus Available Memory) # Target[server.memory]: Free Memory PageTop[server.growright.mempercent]: Free LegendO[server.mempercent]: <H1>Percentage Free Memory</H1> growright.nopercent Unscaled[server.0&memTotalReal.0:craz33guy@localhost ) * 100 / ( memTotalReal.newconns]: Newly Created TCP Connections PageTop[server.cpu]: Used LegendO[server.cpu]: 100 ShortLegend[server.mempercent]: Memory % ShortLegend[server.25Harrison_ch23.newconns]: 10000000000 ShortLegend[server.cpu]: CPU Utilization Legend1[server.G.memory]: <H1>Free Memory</H1> MaxBytes[server.newconns]: tcpPassiveOpens.0:craz33guy@localhost Title[server.P.memory]: gauge.mempercent]: Percentage Free Memory Target[server.mempercent]: ymwd MaxBytes[server.mempercent]: ( memAvailReal.memory]: 100000000000 ShortLegend[server.0:craz33guy@localhost ) options[server.cpu]: growright. in bytes Legend2[server.gauge.memory]: Bytes LegendI[server.

nopercent.0:craz33guy@localhost Title[server.qxd 2/25/05 10:15 AM Page 397 Implementing Advanced Server Monitoring 397 YLegend[server.estabcons]: 10000000000 ShortLegend[server. Run the /usr/bin/mrtg command followed by the name of the configuration file three times.estabcons]: In LegendO[server.disk]: gauge.disk]: <H1>Disk Partition Usage /home and /var</H1> MaxBytes[server.estabcons]: tcpCurrEstab.newconns]: LegendI[server.estabcons]: Established connections Legend2[server.gauge # # Disk Usage Monitoring # Target[server.growright.disk]: ymwd Testing the Configuration The next step is to test that MRTG can load the configuration file correctly.estabcons]: YLegend[server.newconns]: Legend1[server.disk]: 100 ShortLegend[server.disk]: /var Options[server.nopercent. all the files are created and MRTG should operate smoothly.estabcons]: <H1>Established TCP Connections</H1> MaxBytes[server.estabcons]: Options[server.estabcons]: Currently Established TCP Connections PageTop[server.disk]: Disk Partition Usage PageTop[server.25Harrison_ch23. By the third run. [root@bigboy tmp]# service snmpd restart [root@bigboy tmp]# env LANG=C /usr/bin/mrtg /etc/mrtg/server-stats.newconns]: LegendO[server. Restart SNMP to make sure the disk monitoring commands in the snmpd.newconns]: Conns / Min In Out New inbound connections New outbound connections growright.newconns]: Options[server.1&dskPercent.2:craz33guy@localhost Title[server. If all goes well.disk]: dskPercent.estabcons]: growright. MRTG then creates the files.newconns]: Legend2[server. MRTG will complain only about the fact that certain database files don’t exist.cfg .disk]: % YLegend[server.nopercent Unscaled[server.estabcons]: Connections LegendI[server.0&tcpCurrEstab.disk]: Utilization LegendI[server.conf file are activated.estabcons]: Legend1[server.disk]: /home LegendO[server.perminute # # Established TCP Connections # Target[server.

The basic MRTG configuration shown in Chapter 22 provides the data bandwidth statistics. you’ll need to know the exact value of the OID. .cfg Configuring CRON to Use the New MRTG File The final step is to make sure that MRTG is configured to poll your server every five minutes using this new configuration file. other types of systems will need additional MIBs whose correct installation may be unclear in user guides or just not available.d/mrtg file: 0-59/5 * * * * root env LANG=C /usr/bin/mrtg /etc/mrtg/serverstats. [root@bigboy tmp]# indexmaker --output=/var/www/mrtg/index.cfg) plus the new one you created (/etc/mrtg/server-stats.cfg Some versions of Linux require you to edit your /etc/crontab file instead. [root@bigboy tmp]# service crond restart MONITORING NON-LINUX MIB VALUES All the MIBs mentioned so far are for Linux systems. Downloading MIBs from Cisco and using them with the snmpget command was not a success. To do so.cfg /etc/mrtg/server-stats.html \ /etc/mrtg/mrtg. In such cases.qxd 2/25/05 10:15 AM Page 398 398 Advanced MRTG for Linux Chapter 23 Creating a New MRTG Index Page to Include this File Use the indexmaker command and include your original MRTG configuration file from Chapter 22 (/etc/mrtg/mrtg. See Chapter 22 for more details.25Harrison_ch23. You do not know what to do next. You will also have to restart cron with the service crond restart for it to read its new configuration file that tells it to additionally run MRTG every five minutes using the new MRTG configuration file. Scenario Imagine that your small company has purchased a second-hand Cisco switch to connect its Web site servers to the Internet.cfg). as well. but you want to measure the CPU load the traffic is having on the device. add this line to your /etc/cron.

1.109.1. Like most network equipment manufacturers.1) in the original OID file maps to the word “enterprise”.1.1. enter- .1.1.1.1.1.1.6.109.1.109.qxd 2/25/05 10:15 AM Page 399 Monitoring Non-Linux MIB Values 399 Find the OIDs When MIB values fail.4.9.9.9.1.1.1.4.9.1.25Harrison_ch23.6.9.8” Testing The OIDs As you can see.1 SNMPv2-SMI::enterprises.9.3.109.6.9.1 = Gauge32: 33 [root@bigboy tmp]# Success! Now prises. OID value.9.2.1.4.109.1.5.1.2” “1.1.1.1.9.9.5” “1.1.9. The SNMP files for Cisco’s devices can be found at ftp.1.9.3.4.9.1.109. Cisco has an FTP site from which you can download both MIBs and OIDs.4.1.5.1.3.6.1.1.4.1.1.3.1.4” “1.1. Next.3.1.1.1 = Gauge32: 32 SNMPv2-SMI::enterprises.1. The OIDs provided may be incomplete.109.3.1.6.1.1.9.1.6.1.1. OIDs are in the oid directory beneath that.9.3.4.1.3.9.1.1.9.7” “1.8 are not supported.6” “1.1.1.1.1.1.1.109.1.1.6.4. Notice also how SNMP has determined that the first part of the OID value (1.9.9.1.1.1 = Gauge32: 32 [root@bigboy tmp]# Although listed in the OID file.1.1.1.1.oid will contain the necessary values and find these entries inside it: “cpmCPUTotalPhysicalIndex” “cpmCPUTotal5sec” “cpmCPUTotal1min” “cpmCPUTotal5min” “cpmCPUTotal5secRev” “cpmCPUTotal1minRev” “cpmCPUTotal5minRev” “1.1.1.1.109.1.4.1.6.9.1.109.4. you decide that the file CISCO-PROCESSMIB.4.1.1. 1.6. you can use the snmpget command to set only one of the OID values returned by snmpwalk: [root@bigboy tmp]# snmpwalk -v1 -c craz33guy cisco-switch 1.9.1.6.1.1.9.1. so it is best to use the snmpwalk command to try to get all the values below this root first: [root@bigboy tmp]# snmpwalk -v1 -c craz33guy cisco-switch 1.cisco.1 SNMPv2-SMI::enterprises.1.9.9.9.9.1.1.9.3.1..7.1.5.9. all the OIDs are a part of the same tree starting with 1.109. After looking at all the OID files.1.9.1.1.1.109.9.1. and 1.9. you can use this for your MRTG queries. 1.1.1.1 = Gauge32: 32 SNMPv2-SMI::enterprises.1.4.109.6.1.1.1.3” “1.9.6.1.109.1.1.3. it is best to try to find the exact OID value.109.1 = INTEGER: 0 SNMPv2-SMI::enterprises.109.com in the /pub/mibs directory.1.1.1.3.1.3.

4. Check your MRTG parameters to make sure they are correct. CONCLUSION Using the guidelines in this chapter you should be able to graph most SNMP MIB values available on any type of device. These quick steps should be sufficient in most cases and will reward you with a more manageable network. MRTG is an excellent. Verify the IP address and community string of the target device you intend to poll. revise your access controls on the target device and any firewall rules that may impede SNMP traffic. Gauge values defined as counter and vice versa will cause your graphs to have continuous zero values. Graph results that are eight times what you expect may have the bits parameter set.25Harrison_ch23. 3. Ensure you can do an SNMP get of the specific OID value listed in your MRTG configuration file. 2. but because you have done some customizations you’ll have to go the extra mile: 1. . Make sure you can do an SNMP walk of the target device.qxd 2/25/05 10:15 AM Page 400 400 Advanced MRTG for Linux Chapter 23 TROUBLESHOOTING The troubleshooting techniques for advanced MRTG are similar to those mentioned in Chapter 22. flexible monitoring tool and should be considered as a part of any systems administrator’s server management plans. If not.

This makes the correlation of system events on different systems much easier.qxd 2/25/05 10:16 AM Page 401 C H A P T E R 24 The NTP Server IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Download and Install the NTP Package The /etc/ntp. you’ll need a central server that all can access. Sometimes. There are a number of sites that allow the general public to synchronize with them.conf File How to Get NTP Started Testing and Troubleshooting NTP Configuring Cisco Devices to Use an NTP Server Firewalls and NTP Configuring a Windows NTP Client Conclusion The Network Time Protocol (NTP) is a protocol used to help synchronize your Linux system’s clock with an accurate time source.org. not all your servers will have Internet access. It also reduces Internet bandwidth usage due to NTP traffic and reduces the need to manage firewall rules for each NTP client on your network.ntp. consult www. 401 . They are divided into two types: ☞ Stratum 1: NTP sites using an atomic clock for timing ☞ Stratum 2: NTP sites with slightly less accurate time sources It is good practice to have at least one server on your network be the local time server for all your other devices. For a list of available Stratum 1 and 2 servers. in such cases.26Harrison_ch24.

“Installing RPM Software.255. the servers are not allowed to modify the run-time configuration or query your Linux NTP server restrict otherntp.26Harrison_ch24.rpm.server.gov 2. In this example.255.org ntp.gov notrap noquery mask 255.0.server.255 statement is really a subnet mask limiting access to the single IP address of the remote NTP servers.1. then you’ll have to define the networks from which this server will accept NTP synchronization requests. Restrict the type of access you allow these servers.255.0 mask 255.255. 4. Make sure that localhost (the universal IP address used to refer to a Linux server itself) has full access without any restricting keywords: restrict 127.research. 3.conf file is the main configuration file for Linux NTP in which you place the IP addresses of the Stratum 1 and Stratum 2 servers you want to use. as in ntp-4. The syntax is: restrict 192.1 . Downloading and installing RPMs isn’t hard.1.255. THE /ETC/NTP. but it won’t be trusted to be a source of NTP synchronization data.2-5.255.255.0.org # A stratum 2 server at research.0 notrust nomodify notrap In this case the mask statement has been expanded to include all 255 possible IP addresses on the local network. Chapter 6.research. If you need a refresher.org notrap noquery restrict ntp. If this server is also going to provide time for other computers.168. remember that the NTP RPM’s filename usually starts with ntp followed by a version number.gov # A stratum 1 server at server. You do so with a modified restrict statement with the noquery replaced with a notrust keyword.255.CONF FILE The /etc/ntp.255 nomodify mask 255. Here are the steps to create a configuration file using a pair of sample Internet-based NTP servers: 1. other Linux servers.qxd 2/25/05 10:16 AM Page 402 402 The NTP Server Chapter 24 DOWNLOAD AND INSTALL THE NTP PACKAGE Most Red Hat and Fedora Linux software products are available in the RPM format.255 nomodify The mask 255. such as PCs. This allows the network to query your NTP server.” has all the details. Specify the servers you’re interested in: server server otherntp. and networking devices. When searching for the file.i386.

Verifying NTP is Running To test whether the NTP process is running. ☞ To get NTP configured to start at boot. The ntpdate command doesn’t run continuously in the background. Take a look at some sample output of the ntpdate command in which a server whose initial time was set to midnight. Doing an Initial Synchronization If the time on the local server is very different from that of its primary time server. You can now configure other Linux hosts on your network to synchronize with this new master NTP server in a similar fashion. . use the command: [root@bigboy tmp]# pgrep ntpd You should get a response of plain old process ID numbers. Here are some guidelines you can follow to get NTP working correctly. your NTP daemon will eventually terminate itself leaving an error message in the /var/log/messages file. and restart NTP after booting. Save the file and restart NTP for the settings to take effect. Run the ntpdate -u command to force your server to become instantly synchronized with its NTP servers before starting the NTP daemon for the first time. follow these examples: [root@bigboy tmp]# service ntpd start [root@bigboy tmp]# service ntpd stop [root@bigboy tmp]# service ntpd restart TESTING AND TROUBLESHOOTING NTP After configuring and starting NTP. use the line: [root@bigboy tmp]# chkconfig ntpd on ☞ To start. HOW TO GET NTP STARTED You have to restart the NTP process every time you make a change to the configuration file for the changes to take effect on the running process.26Harrison_ch24. you still have to run the ntpd daemon to get continuous NTP updates.qxd 2/25/05 10:16 AM Page 403 Testing and Troubleshooting NTP 403 5. was correctly set to 8:03 a. you should test it to make sure it is working. stop.m.

[root@bigboy tmp]# ntpq -p Here is some sample output of the command: .1.1.168.084943 sec [root@smallfry tmp]# service ntpd start [root@smallfry tmp]# ☞ The date is now corrected: [root@smallfry tmp]# date Thu Aug 12 08:03:45 PDT 2004 [root@smallfry tmp]# Determining If NTP Is Synchronized Properly Use the ntpq command to see the servers with which you are synchronized. run ntpdate and then start ntpd again: [root@smallfry tmp]# service ntpd stop [root@smallfry tmp]# ntpdate -u 192.my-web-site.1.084943 sec [root@smallfry tmp]# ntpdate -u 192.100 offset 0.467652 sec [root@smallfry tmp]# ntpdate -u 192.100 offset 28993. offset.168.1.26Harrison_ch24.168.100 and service ntp host found : bigboy.1. It provided you with a list of configured time servers and the delay.168.1. but it must be run while the ntpd process is stopped.1.qxd 2/25/05 10:16 AM Page 404 404 The NTP Server Chapter 24 ☞ The date was originally set to midnight which was verified by using the date command: [root@smallfry tmp]# date Thu Aug 12 00:00:00 PDT 2004 [root@smallfry tmp]# ☞ The ntpdate command is run three times to synchronize smallfry’s clock to server 192.100 and service ntp host found : bigboy.org 12 Aug 08:03:40 ntpdate[2472]: step time server 192.org 12 Aug 08:03:38 ntpdate[2472]: step time server 192.my-web-site.my-web-site.168.1.168.100. the delay and offset values should be non-zero and the jitter value should be under 100. So you’ll have to stop ntpd.168.168.org 12 Aug 08:03:42 ntpdate[2472]: step time server 192. and jitter that your server is experiencing with them.168.168.100 and service ntp host found : bigboy.100 offset 2. For correct synchronization.100 Looking for host 192.1.100 Looking for host 192.100 Looking for host 192.1.

0 16 u 64 0 0.0.mcs. In these cases.0 16 u .trim 0.1024 0 0.098 3.go ntp0.strat 3 u 507 1024 377 115. remote refid st t when poll reach delay offset jitter ====================================================================== ======= LOCAL(0) LOCAL(0) 10 l 64 7 0.0.000 0.via. 2 u 972 1024 377 38.mcs. In some versions of the Fedora Core 2’s implementation of NTP. This could be located on one of the networks between the NTP server and its time source.000 4000. or firewall software.589 28.018 2.cis.681 -18.424 -3.025 1.wisc.umb.edu ntp1.0.572 milo.000 0. you use the actual IP addresses instead.wisc.00 Your Linux NTP Clients Cannot Synchronize Properly A telltale sign that you haven’t got proper synchronization is when all the remote servers have jitters of 4000 with delay and reach values of 0.000 4000.anl.993 -15.cs.go 2 u 818 1024 125 41.0.392 -mailer1.conectiv.00 nist1.wis ben.anl.strath harris.cs. such as iptables.wisc.549 1.000 0.000 4000.cs.edu 2 u 454 1024 347 54.0 16 u 64 0 0.000 0.edu 2 u 502 1024 357 55.0.psc. clients will not be able to synchronize with a Fedora Core 2 time server unless the notrust nomodify notrap keywords are removed from the NTP client’s restrict statement.127 3.000 0.net .qxd 2/25/05 10:16 AM Page 405 Testing and Troubleshooting NTP 405 remote refid st t when poll reach delay offset jitter ====================================================================== ======== -jj.cs.000 4000.edu gandalf.externa 0.navy.26Harrison_ch24.274 -5. ben.642 *clock.tr 0.GPS.000 0.0 16 u 64 0 0. 1 u 426 1024 377 107.979 0.sigmaso 3 u 95 1024 377 31.c 0. could be running on the server itself. ☞ The notrust nomodify notrap keywords are present in the restrict statement for the NTP client.usno.aol-ca.333 +taylor. .0.00 snvl-smtp1.008 ntp-cup.206 19.534 ntp1. ☞ A firewall blocking access to your Stratum 1 and 2 NTP servers.cc.0.cs.028 -dr-zaius.379 0.047 -ntp0.0.264 1.00 This could be caused by: ☞ Older versions of the NTP package that don’t work correctly if you use the DNS name for the NTP servers.

1.0 Fedora Core 2 File Permissions directory. Instead.26Harrison_ch24. because it is one of the most popular manufacturers of networking equipment and would feature in the overall architectures of many small office/home office (SOHO) environments and corporate departments.16.1.CLIENT NETWORK ------#restrict 172.168. you have to set the group and owner of the directory to be ntp: [root@bigboy tmp]# chown ntp:ntp /etc/ntp If you don’t.100 and 192. you’ll get errors in the /var/log/messages file: Aug 12 00:29:45 smallfry ntpd[2097]: can’t open /etc/ntp/drift. I have included the necessary NTP commands for a variety of Cisco Systems products.255.255.201. the restrict statement has only the client network defined without any keywords and the configuration line that works with other NTP versions has been commented out: # -.0 notrust nomodify notrap restrict 172.1.168.201 ciscorouter(config)# exit ciscorouter# wr mem . Cisco IOS To make your router synchronize with NTP servers with IP addresses 192. in Fedora Core 2.1.255.100 ciscorouter(config)# ntp server 192.TEMP: Permission denied CONFIGURING CISCO DEVICES TO USE AN NTP SERVER You can use NTP to synchronize time on a variety of devices including networking equipment.1. use the commands: ciscorouter> enable password: ********* ciscorouter# config t ciscorouter(config)# ntp update-calendar ciscorouter(config)# ntp server 192.168.255.0 mask 255.0 mask 255.16. the permissions on this directory don’t allow the writing of temporary files.1.qxd 2/25/05 10:16 AM Page 406 406 The NTP Server Chapter 24 In this example. All the Fedora/Red Hat NTP daemons write temporary files to the /etc/ntp Unfortunately.168.

the source port isn’t a high port (above 1023). use the commands: ciscoswitch> enable password: ********* ciscoswitch# set ntp client enable ciscoswitch# ntp server 192. CAT OS To make your router synchronize with NTP servers with IP addresses 192.26Harrison_ch24. but 123 also.” CONFIGURING A WINDOWS NTP CLIENT You can add your new NTP server to your Windows client: 1. Scripts.201 ciscoswitch# exit The ntp server command forms a server association with another system. FIREWALLS AND NTP NTP servers communicate with one another using UDP with a destination port of 123. and Configurations.168.qxd 2/25/05 10:16 AM Page 407 Configuring a Windows NTP Client 407 The ntp server command forms a server association with another system. You’ll have to allow UDP traffic on source/destination port 123 between your server and the Stratum 1/2 server with which you are synchronizing. and enter the name or IP address in the box underneath it. A sample Linux iptables firewall script snippet is in Appendix II. Click on the Update Now button You will get a message saying “Your time has been successfully synchronized” when the operation is complete. “Codes. Click on the time at the bottom right of your screen.100 ciscoswitch# ntp server 192. Click on the Internet Time tab of the dialog box 3.1. Click the check box labeled Automatically Synchronize with an Internet Time Server.1.168.201. Unlike most UDP protocols.1. and set ntp client enable activates the NTP client.168.168. 2.100 and 192.1. . 4. and ntp update-calendar configures the system to update its hardware clock from the software clock at periodic intervals.

The firewall rules can also be greatly simplified. A local NTP server can ensure that the clients all have the same time relative to the server even when Internet connectivity is temporarily lost. Having an NTP server on your local network can make this easier to do. The maintenance of firewall rules for multiple NTP connections to the Internet can also be daunting. even when they all have access there is the risk of them losing synchronization if the central connection to the Internet is lost. and it can also help in the synchronization of time-sensitive transactions. Sometimes it isn’t desirable for all your NTP clients to have access to the Internet to synchronize with Stratum 1 and 2 servers. A local NTP server is frequently a good thing to have for these reasons. especially if the management of the firewall is handled by another group. . thereby reducing the problems of them being out of synchronization with each other.qxd 2/25/05 10:16 AM Page 408 408 The NTP Server Chapter 24 CONCLUSION It is important that all the systems under your control have the same accurate time. It can help to give a very clear indication of a chain of events that involve multiple devices.26Harrison_ch24.

27Harrison_Pt3.qxd 2/25/05 10:17 AM Page 409 P A R T III Advanced Topics 409 .

qxd 2/25/05 10:17 AM Page 410 .27Harrison_Pt3.

using a single installation server (Bigboy) with an IP address of 192. Boot the installation client from a specially created boot CD. HTTP. Connect the new server (installation client) to the same network as the server with the preloaded installation files (installation server). 4. NFS) and the needed network parameters to do this. 3. Enter your preferred installation method (FTP. 2.qxd 2/25/05 10:17 AM Page 411 C H A P T E R 25 Network-Based Linux Installation IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ Setting Up the Installation Server Creating a Boot CD The Network Installation Troubleshooting the Network Installation Differences Between Fedora and Red Hat Installation Automating Installation with Kickstart Conclusion Fedora Linux allows you to do operating system installations via a network connection using a Kickstart server. and the process can be automated.1. Enter your selections. This chapter will briefly explain three methods of installing operating systems over a network connection. 411 . The installation procedure will then continue with the more familiar Fedora Linux installation screens.168. and complete the installation. It is frequently much faster than using CDs.100 as an example. The procedure is fairly simple: 1.28Harrison_ch25.

NFS.28Harrison_ch25. Using a Web server for Kickstart is generally easier because: ☞ Sometimes a Kickstart server has to be located on a remote network. often passing through a firewall. /data/networkand /data/network-install/ISO in which you will copy the necessary files. NFS. Create the directories /data/network-install/Fedora/base. Each method is explained below. This may be important for you when configuring files for automated Kickstart installation. Copy the files from the CD-ROM base directory to the hard disk: [root@bigboy tmp]# cd /mnt/cdrom/Fedora/base [root@bigboy base]# cp -r * /data/network-install/Fedora/base 3. [root@bigboy tmp]# mkdir -p /data/network-install/Fedora/base [root@bigboy tmp]# mkdir -p /data/network-install/Fedora/RPMS [root@bigboy tmp]# mkdir -p /data/network-install/ISO You now need to place the network installation driver files into the base directory. Basic Preparation The first example sets up a Kickstart server for use in Fedora Core 2 installations. You will place all the necessary files in the /data/network-install directory.qxd 2/25/05 10:17 AM Page 412 412 Network-Based Linux Installation Chapter 25 SETTING UP THE INSTALLATION SERVER Kickstart can be configured on an FTP. Create the Installation Directories install/Fedora/RPMS. Strict firewall rules for HTTP are generally easier to configure than those for FTP or NFS. Copying the Files The HTTP. Unmount your CD-ROM. and use the eject command to retrieve it from the drive bay: . Mount your first Fedora CD-ROM: [root@bigboy tmp]# mount /dev/cdrom /mnt/cdrom 2. but my experience is that the Apache server has a number of advantages over the other two. ☞ The http:// nomenclature used by Kickstart for accessing files is more familiar to users than that used for NFS and FTP. or Apache server. and FTP Kickstart methods all require the base set of Fedora files to be installed on the Kickstart server: 1.

and place them in the /data/netdirectory. your /data/network-install/Fedora/ directory should look like this: [root@bigboy tmp]# ls /data/network-install/Fedora/ base RPMS TRANS. When copying is complete.28Harrison_ch25. .. You can download the ISO images from the Fedora Web site or use the Fedora CDs.TBL [root@bigboy tmp]# NFS Preparation work-install/ISO Create ISO images of the installation CDs.. You need about 2GB of space. use: [root@bigboy ISO]# dd if=/dev/cdrom of=FC2-i386-disc2. the code you need is: [root@bigboy tmp]# cd /data/network-install/ISO [root@bigboy ISO]# dd if=/dev/cdrom of=FC2-i386-disc1. make sure they have the same file names as the ones you can download from the Fedora Web site. If you create the ISO files from CDs. use this code to install the fourth CD: [root@bigboy ISO]# dd if=/dev/cdrom of=FC2-i386-disc4.iso bs=32k . HTTP and FTP Preparation Copy all the contents of the /Fedora directory of each of the installation CDs to the /data/network-install/ directory.. For the first CD.iso bs=32k [root@bigboy ISO]# eject cdrom For the third. [root@bigboy ISO]# eject cdrom For the second CD. This requires about 2GB of space as well..iso bs=32k [root@bigboy ISO]# eject cdrom . use: [root@bigboy ISO]# dd if=/dev/cdrom of=FC2-i386-disc3.iso bs=32k [root@bigboy ISO]# eject cdrom Finally.qxd 2/25/05 10:17 AM Page 413 Setting Up the Installation Server 413 [root@bigboy base]# cd /tmp [root@bigboy tmp]# umount /dev/cdrom [root@bigboy tmp]# eject cdrom You also have the option to FTP all the files from the base directory of the desired version of Fedora from the Fedora Web site to the /data/networkinstall/Fedora/base directory.

100> DocumentRoot /data/ </VirtualHost> <Directory /data/network-install> Options +Indexes AllowOverride AuthConfig order allow.1.conf configuration: NameVirtualHost 192.100/network-install/Fedora/RPMS/ or http://192.168.1.100 # # For HTTP Installations # <VirtualHost 192.100/network-install/Fedora/ISO/.1.168.1. Set Up Your FTP Server You also have to set up your VSFTPD server to make incoming anonymous FTP connections log into the /data/network-install directory by default. respectively. Here is a sample httpd.168. You may have to install the mkisofs RPM on newer Fedora versions. [root@bigboy ISO]# mount /mnt/cdrom [root@bigboy ISO]# mkisofs -J -r -T -o filename.168.28Harrison_ch25.iso /mnt/cdrom [root@bigboy ISO]# eject cdrom Set Up Your Web Server You will now have to set up Apache to give the file listings of your /data/network-install/Fedora and /data/network-install/ISO directories by pointing your browser to the URL http://192.deny allow from all </Directory> Remember to restart Apache to make these settings take effect. so don’t forget the mount command. .qxd 2/25/05 10:17 AM Page 414 414 Network-Based Linux Installation Chapter 25 Note Here is a sample procedure to make ISO files with the older mkisofs command. Here is a sample snippet of the vsftpd.conf file: # # Anonymous FTP Root Directory # anon_root=/data/network-install/ # Remember to restart VSFTPD to make these settings take effect. The command requires a mounted CD-ROM drive.

“Remote Disk Access with NFS. and netfs daemons are all running to create an NFS server.sync) 2. Create a /etc/exports file with the following entry in it.my-site. For example: [root@bigboy tmp]# host 192. You must also make sure that the user has read access to the /data/network-install directory. For example: [root@bigboy tmp]# useradd -g users ftpinstall [root@bigboy tmp]# passwd ftpinstall Changing password for user ftpinstall.28Harrison_ch25.in-addr.arpa domain name pointer 192-168-1-96.org has address 192.1.96 [root@bigboy tmp]# This may mean that you must create entries for all your DHCP IP addresses if you choose to use a DHCP method of assigning IP addresses during installation.my-web-site. Make sure that the portmap.” explains this in more detail.myweb-site.96 96.1. not spaces between the entries: /data/network-install *(ro. Chapter 29.qxd 2/25/05 10:17 AM Page 415 Setting Up the Installation Server 415 Create a Special FTP User You can create a special user for non-anonymous FTP installations with the user’s home directory as /. New password: Retype new password: passwd: all authentication tokens updated successfully. nfslock. The startup scripts for these are found in the /etc/init.org. [root@bigboy tmp]# [root@bigboy tmp]# host 192-168-1-96. You must use tabs. nfs. [root@bigboy tmp]# [root@bigbot tmp]# usermod -d / ftpinstall [root@bigbot tmp]# Set Up Your NFS Server The steps for setting up an NFS server are more complicated: 1. a DNS lookup on the IP address of the installation client must return a server name that maps back to the original IP address when a DNS lookup is done on that same server name.192.168. .d directory. Run the exportfs command to add this directory to the NFS database of network available directories: [root@bigboy tmp]# exportfs -ra Keep in mind that the installation client must have a matching pair of forward and reverse DNS entries on your DNS server.168.com 192-168-1-96.1.168. In other words. 3.

the installation procedure mimics the regular Linux installation.iso file onto a CD. 7.100. 6. The remaining menus will be the usual Fedora GUI installation screens. except for the first couple steps: 1.1. Boot your system using the first Fedora installation CD. Go through the usual steps until prompted for the installation method. you now come to the NFS Setup menu. From the list of choices Local CDROM Hard Drive NFS Image FTP HTTP select the network option of your choice (NFS. This would most likely be interface eth0. Connect your client Linux box to the DHCP network.28Harrison_ch25. Press the Enter key. You can either copy an existing disk or burn a copy of the FC2-i386-disc1. “Configuring the DHCP Server. 3. It also reduces the possibility of human error. thereby making it faster. Setting up the Installation Server as a DHCP server is fairly straight forward. . making the Installation client use DHCP during the installation. This is the only CD you’ll need for future network installations. Select DHCP in the Configure TCP/IP screen. For Red Hat Directory specify /data/network-install/ISO. 5. 8.” CREATING A BOOT CD You will need the first CD in your set of Fedora CDs prior to actually starting the installation.qxd 2/25/05 10:17 AM Page 416 416 Network-Based Linux Installation Chapter 25 Configure Your DHCP Server During the installation procedure. Select the Ethernet device with which the installation client is connected to the installation server network. FTP. the installation client prompts you for the IP address it should use for the installation process. I recommend selecting the option that makes the installation client get its address via DHCP. THE NETWORK INSTALLATION From here on. Enter the IP address of the installation server for NFS Server Name. This automates the installation more. Enter the command linux askmethod at the boot: prompt 4.168. for details see Chapter 8. 2. in this case 192. If you selected the NFS method. or HTTP).

. TROUBLESHOOTING THE NETWORK INSTALLATION You can do some basic troubleshooting by accessing the various installation status screens available: ☞ ☞ ☞ ☞ To view the installation logs. If you do select the NonAnonymous FTP box. DIFFERENCES BETWEEN FEDORA AND RED HAT INSTALLATION The RPM file locations have to be different when using the HTTP and FTP methods. which can be useful if you have a large number of servers to install.100. To see kernel messages. Enter the IP address of the installation server when prompted for a Web site name. in this case 192. and the remaining menus will be text-based versions of the usual Fedora installation screens. press Ctrl+Alt+F3 at any time. press Ctrl+Alt+F2. To access a limited BASH shell kernel.cfg Kickstart configuration file.168. Do not add an http:// to the beginning.1. The remaining menus will be text-based versions of the usual Fedora installation screens. press Ctrl+Alt+F4. then the Red Hat directory is /data/network-install. you are given the FTP Setup menu. The remaining menus will be textbased versions of the usual Fedora installation screens. you are shown the HTTP Setup menu. This is the only difference between both operating systems when configuring network installations.1.100 again).28Harrison_ch25. Enter the username and password of your special FTP user account. If you selected the FTP method.168. If you do not select the Non-Anonymous FTP box. the Red Hat directory is simply /. The Red Hat directory is /network-install. You can use this file to create an automated installation of a duplicate system. but Red Hat expects the files to be placed in a directory named RedHat under /data/network-install. Enter the IP address of the installation server as the FTP site name (192. Fedora expects the files to be placed in a directory named Fedora under /data/network-install. AUTOMATING INSTALLATION WITH KICKSTART Both Fedora and Red Hat Linux save all the parameters you used during installation in the /root/anaconda-ks.qxd 2/25/05 10:17 AM Page 417 Automating Installation with Kickstart 417 If you selected the HTTP method. press Ctrl+Alt+F1 from text-based installations or Ctrl+Alt+F7 from the GUI. To return to the main installation screen at any time.

You can disable them by placing a %post section at the end of the Kickstart file with all the post installation commands you wish to run. Fedora comments out the partitioning information in this file.cfg It is possible to use the /root/anaconda-ks. Note Do not change the order of the entries in the Kickstart configuration file. If the server is going to reside on a different network after the installation. adding # signs to comment out certain parameters that may change from system to system.qxd 2/25/05 10:17 AM Page 418 418 Network-Based Linux Installation Chapter 25 This section shows you how to automate network installations using the Kickstart application and NFS. How to Create New Kickstart Configuration Files You can create a customized Kickstart configuration file by using the ksconfig command from a GUI console. It brings up a menu from which you can select all your installation options.28Harrison_ch25. so . then you’ll have to run a separate script to change the IP addressing information after the installation is complete. You can use HTTP and FTP. including the system’s name and IP address. Some processes that Fedora activates by default may not be suitable for your server and may need to be disabled. but I omitted them to keep the discussion brief. Adding post Installation Commands You may want to run some commands on the newly created Linux installation after Kickstart is complete. During the Kickstart process you will be prompted for these unspecified values. You may want to then edit the configuration file.cfg file as a template for future installations. For example: %post chkconfig chkconfig chkconfig chkconfig chkconfig chkconfig isdn off pcmcia off portmap off apmd off nfslock off nfs off A Note about Using anaconda-ks. Note The IP address you assign must be on the same subnet as that of the DHCP server for Kickstart to work. you save the configuration with the filename of your choice. When finished.

verify that the first two lines of the file look like this: install nfs --server=192.168. How to Run a Kickstart Installation work-install directory. you need to issue a command at the boot: prompt to continue with the installation.100:/kickstart/ks.16.1.qxd 2/25/05 10:17 AM Page 419 Automating Installation with Kickstart 419 you have to uncomment it and then make your partitioning modifications or be prepared to be prompted for your portioning information. use: boot: linux ks=nfs:192. Remember that It is best to place your Kickstart files in a subdirectory under the /data/netthe subdirectory is you may want to remove the # comments from the partition section of the file. When using a Web server.168.100/network-install/ If they don’t.100 --dir=/data/network-install/ISO If they don’t you may be prompted for NFS ISO file location information. but the command syntax changes depending on whether you choose NFS or HTTP.168. Booting With Your Kickstart Files There are two ways to specify the name of the Kickstart file to use.1. Manually Specifying the Kickstart Filename After you boot from your boot CD-ROM. The second is to have your DHCP server automatically tell the Kickstart client the name of the Kickstart file when the server assigns the IP address.28Harrison_ch25. use: boot: linux ks=http://192.cfg file is the Kickstart configuration file you want to use.cfg . If not. you will be prompted for this information. When using a NFS server. you may be prompted for RPM base file location information. verify that the first two lines of the file look like this: install url --url http://192. The ks. The first is to enter it manually from the boot: prompt when you insert the boot CD. The upcoming examples assume called /data/network-install/kickstart.100/network-install/kickstart/ks. For the NFS method.1.1.cfg For the HTTP method.

Place your Kickstart file in the /data/network-install/kickstart directory. It then automatically attempts to get a DHCP IP address and see if the DHCP server will specify a configuration file. Kickstart next uses NFS to get both the configuration file and the installation ISOs. you may want to configure your DHCP server to specify the single Kickstart configuration file you wish to use.cfg”. The topic is beyond the scope of this book. 3.168.1.cfg on the boot CD. Insert the boot CD into the Kickstart client Linux box. Edit your dhcpd. filename “/data/network-install/kickstart/ks.100.28Harrison_ch25. . if you have more complex projects. At the boot: prompt type: boot: linux ks Kickstart first searches for a configuration file named ks. A Kickstart server connected to an isolated wireless network dedicated to the purpose may be a good idea for data centers with hundreds of Linux servers. A recent standard called PXE enables you to run Kickstart without a CD-ROM if you configure the NIC to do a network boot from a specially configured DHCP server. The next-server value is the IP address of the Kickstart server. Time is saved not only because a network connection can be faster than using CDs.qxd 2/25/05 10:17 AM Page 420 420 Network-Based Linux Installation Chapter 25 Configuring the Filename Automatically Whenever you have to create lots of cloned Linux servers. next-server 192. but also because it can be left unattended to install a predetermined Linux configuration. 2. and add the following lines to the section for the interface that will be serving DHCP IP addresses. but you may want to research this option some more.conf file. Here is how it’s done: 1. The rest should be automatic. and connect it to the DHCP network. CONCLUSION The Kickstart method of Fedora Linux installation can greatly reduce the length of time it takes to install the operating system.

29Harrison_ch26. A drive failure will corrupt your data.or software-based. 421 . Linear Mode RAID In the Linear RAID. the RAID controller views the RAID set as a chain of disks. Data is written to the next device in the chain only after the previous one is filled. Linear mode RAID is not supported by Fedora Linux. RAID TYPES Whether hardware. RAID can be handled either by the operating system software or it may be implemented via a purpose built RAID disk controller card without having to configure the operating system at all. the chapter focuses on using RAID for partitions that include neither the /boot or the root (/) filesystems. For the sake of simplicity. This chapter will explain how to configure the software RAID schemes supported by Red Hat/Fedora Linux. The aim of Linear RAID is to accommodate large filesystems spread over multiple devices with no data redundancy. RAID can be configured using a variety of standards.qxd 2/25/05 10:18 AM Page 421 C H A P T E R 26 Linux Software RAID IN THIS CHAPTER ☞ ☞ ☞ ☞ RAID Types Before You Start Configuring Software RAID Conclusion The main goals of using redundant arrays of inexpensive disks (RAID) are to improve disk data performance and provide data redundancy. Take a look at the most popular.

1 illustrates the data allocation process in RAID 0. The advantage of RAID 0 is data access speed. RAID 0 can accommodate disks of unequal sizes. the RAID controller tries to evenly distribute data across all disks in the RAID set. RAID 0 is best used with drives of equal size. When this occurs. You should also be aware that RAID 0 is often called striping. Envision a disk as if it were a plate. Like Linear RAID. cherry and strawberry—and four plates. four 9GB hard disks configured in a RAID 0 set are seen by the operating system to be one 36GB disk. because the total number of RAID drives available is reduced. You have four cakes—chocolate. The initialization process of RAID 0 divides the cakes and distributes the slices across all the plates. When RAID runs out of striping space on the smallest device. RAID 0 aims to accommodate large filesystems spread over multiple devices with no data redundancy.qxd 2/25/05 10:18 AM Page 422 422 Linux Software RAID Chapter 26 RAID 0 With RAID 0. RAID 0 is supported by Fedora Linux. A file that is spread over four disks can be read four times as fast.29Harrison_ch26. Block 1 Block 2 Block 3 Block 4 Block 5 Block 6 Block 7 Block 8 Logical Drive Disk Striping (RAID 0) Block 1 Block 3 Block 5 Block 7 Physical Disk #1 Block 1 Block 2 Block 3 Block 4 Block 2 Block 4 Block 6 Block 8 Physical Disk #2 Block 5 Block 6 Block 7 Block 8 Physical Disk #1 Disk Mirroring (RAID 1) Block 1 Block 2 Block 3 Block 4 Block 5 Block 6 Block 7 Block 8 Physical Disk #2 Figure 26.1 RAID 0 and RAID 1 operation. Figure 26. it then continues the striping using the available space on the remaining drives. the data access speed is lower for this portion of data. . For this reason. vanilla. For example. and think of the data as a cake. The RAID 0 drivers make the operating system feel that the cakes are intact and placed on one large plate.

RAID 4 is not supported by Fedora Linux. A limitation of RAID 1 is that the total RAID size in gigabytes is equal to that of the smallest disk in the RAID set. When the failed disk is replaced. RAID 4 requires at least three disks in the RAID set and can survive the loss of a single drive only. Figure 26. The parity disk can therefore act as a bottleneck. it is repopulated with the lost data with the help of the parity disk’s information. RAID 1 offers data redundancy. This RAID method is therefore frequently called disk mirroring. With a hardware-based solution. Think of telling two people the same story so that if one forgets some of the details you can ask the other one to remind you. without the speed advantages of RAID 0. and the disk controller then duplicates the data to the mirror disks. RAID 5 can survive the loss of a single disk only. any data written to any section of the data portion of the RAID set must be followed by an update of the parity disk. . When one of the disks in the RAID set fails. RAID 4 combines the high speed provided of RAID 0 with the redundancy of RAID 1.1 illustrates the data allocation process in RAID 1. When the failed disk is replaced. while maintaining many of the speed features of RAID 0 and the redundancy of RAID 1. This avoids the parity disk bottleneck. data is cloned on a duplicate disk. This can saturate data busses and CPU use. In other words. RAID 4 isn’t used very frequently. RAID 1 also offers the possibility of using a hot standby spare disk that will be automatically cloned in the event of a disk failure on any of the primary RAID devices. but the parity information is not. A disadvantage of software-based RAID 1 is that the server has to send data twice to be written to each of the mirror disks. Its major disadvantage is that the data is striped. the other one continues to function. For this reason. RAID 4 RAID 4 operates like RAID 0 but inserts a special error-correcting or parity chunk on an additional disk dedicated to this purpose.qxd 2/25/05 10:18 AM Page 423 RAID Types 423 RAID 1 With RAID 1. When this occurs. RAID 5 RAID 5 improves on RAID 4 by striping the parity data between all the disks in the RAID set. the data in it can be recreated on the fly with the aid of the information on the RAID set’s parity disk. Like RAID 4. the server CPU sends the data to the RAID disk controller once. the data is automatically cloned to the new disk from the surviving disk. RAID 1 is supported by Fedora Linux.29Harrison_ch26. This makes RAID-capable disk controllers the preferred solution when implementing RAID 1. the extra space on the larger device isn’t used. Unlike RAID 0.

Block 1 Block 2 Block 3 Block 1 Block 4 Block 5 Block 6 Block 6 Block 7 Block 8 Logical Drive Block 7 Block 8 Error Check Blocks 7 and 8 Physical Disk #3 Error Check Blocks 3 and 4 Block 3 Block 4 Block 2 Error Check Blocks 1 and 2 Error Check Blocks 5 and 6 Block 5 Physical Disk #1 Physical Disk #2 RAID 5 Figure 26. IDE Drives To save costs.qxd 2/25/05 10:18 AM Page 424 424 Linux Software RAID Chapter 26 RAID 5 is supported by Fedora Linux.2 RAID 5 operation. so you can configure them right after your system’s power on self test (POST). the hardware does all the work. Linux RAID 5 requires a minimum of three disks or partitions. many small business systems will probably use IDE disks.29Harrison_ch26. Hardware-based RAID is transparent to your operating system.2 illustrates the data allocation process in RAID 5. then you should be aware of these basic guidelines to follow when setting up software RAID. They usually have their own BIOS. If hardware RAID isn’t available. BEFORE YOU START Specially built hardware-based RAID disk controllers are available for both IDE and SCSI drives. Figure 26. but they do have some limitations: .

☞ There are no jumpers to set on Serial ATA drives to make it a master or slave. You cannot replace them while your system is running.29Harrison_ch26. If you can’t afford more expensive and faster SCSI drives. ☞ The failure of one drive on an IDE bus often causes the malfunctioning of the second device. This can be fatal if you have two IDE drives of the same RAID set attached to the same cable. I recommend you use only one IDE drive per controller when using RAID. drives as the preferred entry-level disk storage option because of a number of advantages: ☞ The drive data cable can be as long as one meter in length versus IDE’s 18 inches. SCSI Drives SCSI hard disks have a number of features that make them more attractive for RAID use than either IDE or Serial ATA drives: ☞ SCSI controllers are more tolerant of disk failures. ☞ Only two devices can be attached per controller.qxd 2/25/05 10:18 AM Page 425 Before You Start 425 ☞ The total length of an IDE cable can be only a few feet long. which makes them easier to configure. Serial ATA Drives Serial ATA type drives are rapidly replacing IDE. ☞ There is only one drive per cable. ☞ Serial ATA has better error checking than IDE. For these reasons. or ultra ATA. . ☞ The performance of the IDE bus can be degraded by the presence of a second device on the cable. In a home or SOHO setting. The failure of a single drive is less likely to disrupt the remaining drives on the bus. IDE-based software RAID may be adequate. possible without the fear of affecting other devices on the data cable. especially in a corporate environment. ☞ IDE drives do not hot swap. ☞ IDE drives have a 133Mbytes/s data rate whereas the Serial ATA specification starts at 150 Mbytes/sec with a goal of reaching 600 Mbytes/s over the expected ten year life of the specification. which makes hot swapping or the ability to replace components while the system is still running. Serial ATA would be the preferred device for software and hardware RAID. which generally limits IDE drives to small home systems.

If you decide to use RAID. which allows you to replace them while the system is running.qxd 2/25/05 10:18 AM Page 426 426 Linux Software RAID Chapter 26 ☞ SCSI cables can be up to 25 meters long. however. A mistake could unintentionally corrupt valid data. issue the exit command. ☞ Much more than two devices may be connected to a SCSI cable bus. It can accommodate 7 (single-ended SCSI) or 15 (all other SCSI types) devices. causes all the old data on the underlying RAID partitions to be lost. You have to format this new RAID device before your Linux system can store files on it. Configure RAID in Single-User Mode As you will be modifying the disk structure of your system. which may make them less attractive for home use. It is best to back up the data on these and any other partitions on the disk drive on which you want implement RAID. Should I Use Software RAID Partitions or Entire Disks? It is generally not a good idea to share RAID-configured partitions with nonRAID partitions. [root@bigboy tmp]# init 1 Once finished. Back up Your System First Software RAID creates the equivalent of a single RAID virtual disk drive made up of all the underlying regular partitions used to create it. and your system will boot in the default runlevel provided in the /etc/inittab file. SCSI drives tend to be more expensive than IDE drives.29Harrison_ch26. all the partitions on each RAID disk should be part of a RAID set. Many people simplify this problem by filling each disk of a RAID set with only one partition. ☞ SCSI currently supports data rates of up to 640 Mbytes/s making them highly desirable for installations where rapid data access is imperative. This makes sure that most applications and networking are shutdown and that no other users can access the system. ☞ Some models of SCSI devices support hot swapping. . reducing the risk of data corruption during the exercise. making them suitable for data center applications. Formatting. however. The reason for this is obvious: A disk failure could still incapacitate a system. you should also consider configuring RAID while your system is running in single-user mode from the VGA console.

Here is some sample output of these commands: [root@bigboy tmp]# fdisk –l Disk /dev/hda: 12.. Determining Available Partitions First use the fdisk -l command to view all the mounted and unmounted filesystems available on your system. which shows only mounted filesystems but has the big advantage of giving you the mount points too. . If you are doing RAID 0 or RAID 5. You may then also want to use the df -k command. The partitions to be used are: ☞ ☞ ☞ /dev/hde1 /dev/hdf2 /dev/hdg1 Be sure to adapt the various stages outlined below to your particular environment. 63 sectors/track. In the tutorial example. RAID Partitioning You first need to identify two or more partitions. the partitions should be of approximately the same size.. you’ll be configuring RAID 5 using a system with three pre-partitioned hard disks. each on a separate disk.0 GB.. RAID limits the extent of data access on each partition to an area no larger than that of the smallest partition in the RAID set..29Harrison_ch26.qxd 2/25/05 10:18 AM Page 427 Configuring Software RAID 427 CONFIGURING SOFTWARE RAID Configuring RAID using Fedora Linux requires a number of steps that need to be followed carefully. 12072517632 bytes 255 heads. /dev/hda15 1455 [root@bigboy tmp]# End 13 144 209 1467 655 Blocks 104391 1052257+ 522112+ 10104885 3582463+ Id 83 83 82 5 83 System Linux Linux Linux swap Extended Linux 1467 104391 83 Linux . as in this scenario. 1467 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start /dev/hda1 * 1 /dev/hda2 14 /dev/hda3 145 /dev/hda4 210 /dev/hda5 210 . These two commands should help you to easily identify the partitions you want to use.

g..... old versions of LILO) 2) booting and partitioning software from other OSs (e. so you need to make sure they are unmounted: [root@bigboy tmp]# umount /dev/hde1 [root@bigboy tmp]# umount /dev/hdf2 [root@bigboy tmp]# umount /dev/hdg1 Prepare the Partitions with fdisk You have to change each partition in the RAID set to be of type FD (Linux raid autodetect).. p q s t ..29Harrison_ch26..qxd 2/25/05 10:18 AM Page 428 428 Linux Software RAID Chapter 26 [root@bigboy tmp]# df –k Filesystem Mounted on /dev/hda2 /dev/hda1 /dev/hda15 .. print the partition table quit without saving changes create a new empty Sun disklabel change a partition’s system id Command (m for help): .. Here is an example using /dev/hde1: [root@bigboy tmp]# fdisk /dev/hde The number of cylinders for this disk is set to 8355. /dev/hda7 [root@bigboy tmp]# 1K-blocks 1035692 101086 101086 163916 8357 4127 Used Available Use% 819164 87510 91740 17% / 9% /boot 5% /data1 5336664 464228 4601344 10% /var Unmount the Partitions You don’t want anyone else accessing these partitions while you are creating the RAID set. . .. . but this is larger than 1024. .. DOS FDISK... and you can do this with fdisk..g. There is nothing wrong with that. and could in certain setups cause problems with: 1) software that runs at boot time (e... OS/2 FDISK) Command (m for help): Use fdisk Help Now use the fdisk m command to get some help: Command (m for help): m .

Modify its type using the t command.. 4311982080 bytes 16 heads. 16 Hidden FAT16 17 Hidden HPFS/NTF 18 AST SmartSleep 1b Hidden Win95 FA Hex code (type L to Changed system type 61 SpeedStor a9 NetBSD f2 DOS secondary 63 GNU HURD or Sys ab Darwin boot fd Linux raid auto 64 Novell Netware b7 BSDI fs fe LANstep 65 Novell Netware b8 BSDI swap ff BBT list codes): fd of partition 1 to fd (Linux raid autodetect) Command (m for help): Make Sure the Change Occurred Use the p command to get the new proposed partition table: Command (m for help): p Disk /dev/hde: 4311 MB. 8355 cylinders Units = cylinders of 1008 * 512 = 516096 bytes Device Boot /dev/hde1 /dev/hde2 /dev/hde4 /dev/hde5 /dev/hde6 Start 1 4089 6608 6608 7501 End 4088 5713 8355 7500 8355 Blocks 2060320+ 819000 880992 450040+ 430888+ Id fd 83 5 83 83 System Linux raid autodetect Linux Extended Linux Linux Command (m for help): Save the Changes Use the w command to permanently save the changes to disk /dev/hde: Command (m for help): w The partition table has been altered! Calling ioctl() to re-read partition table. . 63 sectors/track.29Harrison_ch26..qxd 2/25/05 10:18 AM Page 429 Configuring Software RAID 429 Set the ID Type to FD Partition /dev/hde1 is the first partition on disk /dev/hde... and specify the partition number and type code.. .. . You also should use the L command to get a full listing of ID types in case you forget: Command (m for help): t Partition number (1-5): 1 Hex code (type L to list codes): L .

1. For example. they would be numbered 0. I won’t show the process for the other partitions. The data is distributed across the drives in 32MB chunks: # # sample raiddev configuration file # ‘old’ RAID0 array created with mdtools. Repeat for the Other Partitions For the sake of brevity. and the set of three is called /dev/md0. # raiddev /dev/md0 raid-level 5 nr-raid-disks 3 persistent-superblock 1 chunk-size 32 parity-algorithm left-symmetric device /dev/hde1 raid-disk 0 device /dev/hdf2 raid-disk 1 device /dev/hdg1 raid-disk 2 . You can find templates for this file in the /usr/share/doc/raidtools* directory. persistent-superblock must be set to 0. 2.29Harrison_ch26. and 3. the /etc/raidtab persistent-superblock must be set to 1 for the RAID autodetect feature (partition type FD) to work. RAID 5 is configured to use each of the desired partitions on the three disks. For an explanation of the various parameters. remember these general guidelines: ☞ When configuring RAID 5. It’s enough to know that the steps for changing the IDs for /dev/hdf2 and /dev/hdg1 are very similar. [root@bigboy tmp]# The error above will occur if any of the other partitions on the disk is mounted. To ensure success. if you have four partitions for RAID.qxd 2/25/05 10:18 AM Page 430 430 Linux Software RAID Chapter 26 WARNING: Re-reading the partition table failed with error 16: Device or resource busy. Edit the RAID Configuration File The Linux RAID configuration file is /etc/raidtab. The kernel still uses the old table. you must use a parity-algorithm setting. Here. ☞ The raid-disk parameters for each partition in the /etc/raidtab file are numbered starting at 0. issue the man raidtab command. 4. The new table will be used at the next reboot. For all other RAID versions. Consider an example. and 5. ☞ For RAID levels 1. Syncing disks.

The next example uses the -j qualifier to ensure that a journaling filesystem is created. which is comprised of 8 blocks. Confirm that the initialization is finished by inspecting the file and making sure that there are no initialization-related messages: [root@bigboy tmp]# cat /proc/mdstat Personalities : [raid5] read_ahead 1024 sectors md0 : active raid5 hdg1[2] hde1[1] hdf2[0] 4120448 blocks level 5. algorithm 3 [3/3] [UUU] unused devices: <none> [root@bigboy tmp]# Format the New RAID Set Your new RAID device now has to be formatted. 104391kB. raid superblock at 104320kB disk 2: /dev/hdg1. 1030160 blocks 51508 blocks (5. you will get parity errors. raid superblock at 104320kB disk 1: /dev/hdf2. raid superblock at 104320kB [root@bigboy tmp]# Confirm RAID Is Correctly Initialized The /proc/mdstat file provides the current status of all RAID devices. Here a block size of 4KB (4096 bytes) is used with each chunk. [root@bigboy tmp]# mkraid /dev/md0 analyzing super-block disk 0: /dev/hde1. It is very important that the chunk-size parameter in the /etc/raidtab file match the value of the block size multiplied by the stride value in the command below.qxd 2/25/05 10:18 AM Page 431 Configuring Software RAID 431 Create the RAID Set The mkraid command creates the RAID set by reading the The example creates the logical RAID device /dev/md0: /etc/raidtab file. 104391kB. 32k chunk.32 (09-Nov-2002) Filesystem label= OS type: Linux Block size=4096 (log=2) Fragment size=4096 (log=2) 516096 inodes.29Harrison_ch26. 104391kB. If the values don’t match.00%) reserved for the super user First data block=0 32 block groups 32768 blocks per group. 32768 fragments per group 16128 inodes per group . [root@bigboy tmp]# mke2fs -j -b 4096 -R stride=8 /dev/md0 mke2fs 1.

d/rc. 294912. make the Linux operating system fully aware of the RAID set by loading the driver for the new RAID set using the raidstart command: [root@bigboy tmp]# raidstart /dev/md0 [root@bigboy tmp]# Create a Mount Point for the RAID Set After the driver loads. 229376. Mounting a RAID device that doesn’t have a loaded driver can corrupt your data and produce this error: Starting up RAID devices: md0(skipped) Checking filesystems . the /dev/md0 device: /dev/md0 /mnt/raid ext3 defaults 1 2 Do not use labels in the /etc/fstab file for RAID devices. 98304. Device mounting then occurs later on in the boot process. just use the real device name. whichever comes first. create a mount point for called /mnt/raid: [root@bigboy mnt]# mkdir /mnt/raid /dev/md0. such as /dev/md0. 884736 Writing inode tables: done Creating journal (8192 blocks): done Writing superblocks and filesystem accounting information: done This filesystem will be automatically checked every 26 mounts or 180 days. The script will not automatically start the RAID set driver for the RAID set if it doesn’t find a match. the /etc/rc. such as this one Edit the /etc/fstab File The /etc/fstab file lists all the partitions that need to mount when the system boots. 819200. On startup. Add an Entry for the RAID set.29Harrison_ch26.sysinit script checks the /etc/fstab file for device entries that match RAID set names in the /etc/raidtab file. [root@bigboy tmp]# Load the RAID Driver for the New RAID Set Next. Use tune2fs -c or -i to override.qxd 2/25/05 10:18 AM Page 432 432 Linux Software RAID Chapter 26 Superblock backups stored on blocks: 32768. 163840.

/dev/hdf2. use the man fstab command to get a comprehensive explanation of each data column it contains. ***journal has been deleted . Run fsck manually (ie without -a or -p options). If you are not familiar with the /etc/fstab file. Either the superblock or the partition table is likely to be corrupt! /boot: clean.29Harrison_ch26. You have your choice of methods: ☞ The mount command’s -a flag causes Linux to mount all the devices in the /etc/fstab file that have automounting enabled (default) and that are also not already mounted. and /dev/hdg1 partitions were replaced by the combined /dev/md0 partition. 41/26104 files. [root@bigboy tmp]# mount -a ☞ You can also mount the device manually.file system is now ext 2 only*** /raiddata: The filesystem size (according to the superblock) is 2688072 blocks. [root@bigboy tmp]# raidstart /dev/md0 Mount the New RAID Set Use the mount command to mount the RAID set.qxd 2/25/05 10:18 AM Page 433 Configuring Software RAID 433 /raiddata: Superblock has a bad ext3 journal(inode8) CLEARED. Make sure that all references to them in this file are commented with a # at the beginning of the line or deleted entirely: #/dev/hde1 #/dev/hdf2 #/dev/hdg1 /data1 /data2 /data3 ext3 ext3 ext3 defaults defaults defaults 1 2 1 2 1 2 Start the New RAID Set’s Driver You now can start the new RAID set’s driver with the raidstart command. 12755/104391 blocks /raiddata: UNEXPECTED INCONSISTENCY. The /dev/hde1. [root@bigboy tmp]# mount /dev/md0 /mnt/raid . so you’ll only have to do this once. This command is run automatically at boot time. You therefore don’t want the old partitions to be mounted again. The physical size of the device is 8960245 blocks.

Remember to take these factors into consideration when determining the right solution for your needs and research the topic thoroughly before proceeding. 32k chunk. it is absolutely transparent to Linux.29Harrison_ch26. is provided after you load the driver using the raidstart command. the file has very little information. but it tends to be slower and less reliable than RAID provided by a hardwarebased RAID disk controller. Weighing cost versus reliability is always a difficult choice in systems administration. . Hardware RAID configuration is usually done via the system BIOS when the server boots up. and once configured. [root@bigboy tmp]# raidstart /dev/md0 [root@bigboy tmp]# cat /proc/mdstat Personalities : [raid5] read_ahead 1024 sectors md0 : active raid5 hdg1[2] hde1[1] hdf2[0] 4120448 blocks level 5. When the RAID driver is stopped.qxd 2/25/05 10:18 AM Page 434 434 Linux Software RAID Chapter 26 Check the Status of the New RAID The /proc/mdstat file provides the current status of all the devices. hardware RAID requires entire disks to be dedicated to the purpose and when combined with the fact that it usually requires faster SCSI hard disks and an additional controller card. algorithm 3 [3/3] [UUU] unused devices: <none> [root@bigboy tmp]# CONCLUSION Linux software RAID provides redundancy across partitions and hard disks. Unlike software RAID. including the partitions of the RAID set. as seen here: [root@bigboy tmp]# raidstop /dev/md0 [root@bigboy tmp]# cat /proc/mdstat Personalities : [raid5] read_ahead 1024 sectors unused devices: <none> [root@bigboy tmp]# More information. it tends to be expensive.

The first is by moving directories from a full partition to an empty one made available by the new disk and then linking the directory structures of the two disks together. the /var partition would still be over 95% full. Scenario Things are getting crowded on bigboy: Even after you removed all unwanted data. or maybe you ran out of available space. To provide more space. Perhaps an existing device failed. The second is by merging the partitions together to create a combined partition using the Linux Logical Volume Manager (LVM). This chapter explores how to add a disk to a Linux system in two ways. You need to add a new 435 . ADDING DISKS TO LINUX At some stage you’ll be faced with the task of installing an additional hard drive into your Linux server. The most common reasons for this are expanding databases. increasing numbers of users. this section will cover adding a hard disk with only one partition and will then explain how to migrate data from the full disk to the new one. and the larger number of tasks your Linux server is expected to perform until a replacement is found.qxd 2/25/05 10:18 AM Page 435 C H A P T E R 27 Expanding Disk Capacity IN THIS CHAPTER ☞ Adding Disks to Linux ☞ Expanding Partitions with LVM ☞ Conclusion The lack of available disk storage frequently plagues Linux systems administrators.30Harrison_ch27.

. the first disk would be a. In the example. and so on. Using this convention the fifth partition on the fourth IDE drive would be /dev/hdd5. SCSI disks all start with sd and IDE disks with hd. After this comes a letter that identifies the unit number of the disk. Determining the Disk Types Linux stores the names of all known disk partitions in the /proc/partitions file.qxd 2/25/05 10:18 AM Page 436 436 Expanding Disk Capacity Chapter 27 hard drive to the system. the system has two hard disks. The entire hard disk is represented by an entry with a minor number of 0. a two-digit number defines the partition number. the second would be b. but you now need to know how to proceed. disk /dev/hda has been partitioned. but the new disk (/dev/hdb) needs to be prepared to accept data. [root@bigboy tmp]# cat /proc/partitions major minor #blocks name 3 0 7334145 3 1 104391 3 2 1052257 3 3 2040255 3 4 1 3 5 3582463 3 6 554211 22 0 78150744 [root@bigboy tmp]# hda hda1 hda2 hda3 hda4 hda5 hda6 hdb Note Linux hard disk device names follow a specific standard. Finally.30Harrison_ch27. You can verify this situation with the df -k command’s output. which also shows that the other partitions are too full to accept any more data: [root@bigboy tmp]# df -k Filesystem 1K-blocks /dev/hda3 505636 /dev/hda1 101089 none 63028 /dev/hda5 248895 /dev/hda7 3304768 /dev/hda2 3304768 [root@bigboy tmp]# Used Available Use% Mounted on 118224 361307 25% / 14281 81589 15% /boot 0 63028 0% /dev/shm 6613 229432 3% /tmp 2720332 416560 87% /usr 3300536 4232 99% /var A new hard disk was added according to the manufacturer’s instructions. the third would be c. and all the partitions on the drive are sequentially numbered after that. so for example.

63 sectors/track. You want to run fdisk on the /dev/hdb disk. Just to make sure you’re on the correct device. 80026361856 bytes . and could in certain setups cause problems with: 1) software that runs at boot time (e..30Harrison_ch27. In this case. default 9729): 4.g.qxd 2/25/05 10:18 AM Page 437 Adding Disks to Linux 437 Preparing Partitions on New Disks Linux partition preparation is very similar to that in a Windows environment. Run the print (p) command to confirm that you successfully created the partition.0 GB. OS/2 FDISK) Command (m for help): 2. but this is larger than 1024. Type the fdisk command followed by the name of the disk. 9729 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start End Blocks Id System Command (m for help): 3. The first Linux step in adding a new disk is to partition it in preparation of adding a filesystem to it. and use the defaults to make the partition occupy the entire disk: Command (m for help): n Command action e extended p primary partition (1-4) p Partition number (1-4): 1 First cylinder (1-9729. there are none. 80026361856 bytes 255 heads. issue the p command to print all the known partitions on the disk. You will see that n is the command to add a new partition. old versions of LILO) 2) booting and partitioning software from other OSs (e. Command (m for help): p Disk /dev/hdb: 80. which is good: Command (m for help): p Disk /dev/hdb: 80. The fdisk m command prints a small help manual of valid commands. number 1..0 GB. There is nothing wrong with that. because both operating systems share the fdisk partitioning utility. The steps are: 1. Add a new primary partition.g. DOS FDISK. default 1):<RETURN> Using default value 1 Last cylinder or +size or +sizeM or +sizeK (1-9729. so the command is [root@bigboy tmp]# fdisk /dev/hdb The number of cylinders for this disk is set to 9729.

you can use the d command to delete the partition and start over. or save. this won’t be necessary. The t command enables you to change partition type from the default of 83 for regular Linux partitions to something else. 913 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot /dev/hda1 * /dev/hda2 /dev/hda3 /dev/hda4 /dev/hda5 /dev/hda6 Start 1 14 145 399 399 845 End 13 144 398 913 844 913 Blocks 104391 1052257+ 2040255 4136737+ 3582463+ 554211 Id 83 83 82 5 83 83 System Linux Linux Linux swap Extended Linux Linux Command (m for help): Adding more partitions is just a question of repeating the previous steps the required number of times. 63 sectors/track. 7510164480 bytes 255 heads. 63 sectors/track. exit with the q command.30Harrison_ch27. Here is an example of a partition table that includes an extended partition followed by two regular partitions within it: Command (m for help): p Disk /dev/hda: 7510 MB. the default value is sufficient. Do that now. 5. Note When you created the new partition. you may have noticed that fdisk queried you as to whether it was going to be a primary or secondary partition. if you need more. Command (m for help): w Command (m for help): q .qxd 2/25/05 10:18 AM Page 438 438 Expanding Disk Capacity Chapter 27 255 heads. Changes won’t be made to the disk’s partition table until you use the w command to write. In most cases. the changes. you can convert one of the primary ones into an extended one. you may need to add an extended partition. Linux allows only four primary partitions. while remembering that at some stage. and. such as 82 for swap space. when finished. 9729 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot /dev/hdb1 Start 1 End 9726 Blocks 78148161 Id 83 System Linux Command (m for help): Tip If you make a mistake.

. Create directory /mnt/hdb1 for this purpose: [root@bigboy tmp]# mkdir /mnt/hdb1 .. . Disk /dev/hdb: 80.30Harrison_ch27. The Fedora installation procedure defaults to an ext3 type.. giving it a new directory structure by using the mkfs command. . .. .. which is what you should use here: [root@bigboy tmp]# mkfs -t ext3 /dev/hdb1 Next. These steps will be covered next...qxd 2/25/05 10:18 AM Page 439 Adding Disks to Linux 439 After this is complete. 22 0 22 1 [root@bigboy tmp]# cat /proc/partitions #blocks name -l command 78150744 hdb 78150744 hdb1 tmp]# [root@bigboy tmp]# fdisk -l . 63 sectors/track.. 80026361856 bytes 255 heads..0 GB. Verifying Your New Partition You can take a look at the /proc/partitions file or use the fdisk to see the changes to the disk partition structure of your system: [root@bigboy major minor ... you’ll need to verify your work and start migrating your data to the new disk. 9729 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot /dev/hdb1 [root@bigboy tmp]# Start 1 End 9729 Blocks 76051710 Id 83 System Linux Putting a Directory Structure on Your New Partition You now need to format the partition. you must create a special mount point directory.. to which the new partition will be attached.

The root (or master) filesystem has a value of 1 and is checked on the first pass. The fifth entry governs whether the dump filesystem backup command can be used for the filesystem. You don’t have to wait for a reboot to mount your partition. [root@bigboy tmp]# df -k Filesystem 1K-blocks /dev/hda3 505636 /dev/hda1 101089 none 63028 /dev/hda5 248895 /dev/hda7 3304768 /dev/hda2 3304768 [root@bigboy tmp]# df -k command.qxd 2/25/05 10:18 AM Page 440 440 Expanding Disk Capacity Chapter 27 When Linux boots. If you are not familiar with the /etc/fstab file use the man fstab command to get a full explanation of its various options.30Harrison_ch27. The final entry defines the order in which a filesystem check is done at boot time. You’ll have to add an entry for your new partition that looks like this: # # File: /etc/fstab # /dev/hdb1 /mnt/hdb1 ext3 default 1 2 The first entry is the name of the partition followed by the mount point directory and the filesystem type. which need be only default for most scenarios. You can use the mount command with the -a option to read the /etc/fstab file for new entries: [root@bigboy tmp]# mount -a You are now able to access your new partition as device /mnt/hdb1. A value of 0 means no. and 1 means yes. Migrating Data to Your New Partition As you remember from investigating with the tion is almost full. it searches the /etc/fstab file for a list of all partitions and their mounting characteristics. You can recursively use the command by using the cd command to step down through all the subdirectories until you discover the one with the . all other filesystems should have a value of 2. The fourth entry defines the mounting options. the /var parti- Used Available Use% Mounted on 118224 361307 25% / 14281 81589 15% /boot 0 63028 0% /dev/shm 6613 229432 3% /tmp 2720332 416560 87% /usr 3300536 4232 99% /var The du -sk * command shows the disk usage of all subdirectories in a directory. and then it mounts the partitions automatically. The check is done twice.

. use these steps: 1. If other users are present.my-web-site. empty mount point: /var/transactions directory.05b# umount /mnt/hdb1 .. this will later act as a sh-2.05b# mkdir /var/transactions 6. send a message with the wall command informing them that the system is about to shutdown.. [root@bigboy tmp]# who root pts/0 Nov bob pts/0 Nov bunny pts/0 Nov [root@bigboy tmp]# wall The 6 14:46 (192-168-1-242.org) 6 16:25 (192-168-1-250. you only had to go to the that the /var/transactions directory was the culprit.. Use the who command to see who’s logged in.05b# cp -r /var/transactions-save /mnt/hdb1 7. In this case.. Rename the /var/transactions directory /var/transactions-save to make sure you have an easy to restore backup of the data. 133784 transactions .05b# mv /var/transactions /var/transactions-save 5.my-web-site. Create a new.. . 2. [root@bigboy tmp]# cd /var [root@bigboy var]# du -sk * 2036 cache 4 db 8 empty . Copy the contents of the /var/transactions-save directory to the root directory of /dev/hdb1. Log into the VGA console. not just the tapes: sh-2.. [root@bigboy var]# /var directory to see As a solution.org) system is shutting down now! 7 15:04:27 2004): Broadcast message from root (pts/0) (Sun Nov The system is shutting down now! [root@bigboy tmp]# 3. Unmount the new /dev/hdb1 partition: sh-2. and enter single-user mode: [root@bigboy tmp]# init 1 4. which is actually /mnt/hdb1: sh-2.qxd 2/25/05 10:18 AM Page 441 Adding Disks to Linux 441 greatest file usage. . the /var partition will be expanded to the new /dev/hdb1 partition mounted on the /mnt/hdb1 directory mount point. Back up the data on the partition you are about to work on..my-web-site. To migrate the data.30Harrison_ch27.org) 6 12:01 (192-168-1-248.

you need to become familiar with some LVM concepts: ☞ Physical Volume: A physical volume (PV) is another name for a regular physical disk partition that is used or will be used by LVM. Linux also allows you to merge partitions together in order to create a larger combined one. The reasons and steps for doing so will be explained next.05b# exit 12. Remount /dev/hdb1 on the new mount point using the mount -a command. 11. # # File: /etc/fstab # #/dev/hdb1 /dev/hdb1 /mnt/hdb1 ext3 default 1 2 ext3 default 1 2 /var/transactions 9. . EXPANDING PARTITIONS WITH LVM The Logical Volume Manager (LVM) enables you to resize your partitions without having to modify the partition tables on your hard disk.30Harrison_ch27. removing our previous entry for /dev/hdb1 replacing it with one using the new mount point. Make sure your applications are working correctly and delete both the /var/transactions-save directory and the /mnt/hdb1 mount point directory at some later date.05b# mount -a 10. Return to multi-user mode by typing exit. This is most useful when you find yourself running out of space on a filesystem and want to expand into a new disk partition versus migrating all or a part of the filesystem to a new disk. The system will return to its default runlevel. which reads /etc/fstab and automatically mounts any entries that are not mounted already: sh-2. Edit the /etc/fstab file. Test to make sure that the contents of the new /var/transactions directory is identical to /var/transactions-save. This exercise showed you how to migrate the entire contents of a subdirectory to a new disk. LVM Terms Before you run through the LVM tutorial. sh-2.qxd 2/25/05 10:18 AM Page 442 442 Expanding Disk Capacity Chapter 27 8.

☞ Physical Extent: Real disk partitions are divided into chunks of data called physical extents (PEs) when you add them to a logical volume. Make sure you understand these terms fully as they will be used repeatedly in many of the following sections. but there isn’t the budget to purchase an adequately sized hard drive. The LVM process will destroy the data on all physical volumes. volume groups are analogous to a virtual disk drive. If other users are present. You just added a new hard drive /dev/hdf with 50% of the capacity of /dev/hde5 into which you want to expand /home. Under LVM.org) (192-168-1-250. Suppose a small company needs to expand disk capacity. Let’s go! Configuring LVM Devices It is probably best to learn about the features of LVM through a scenario. Unmount Your /home Filesystem As /home stores most users’ data. This may seem complicated.org) is shutting down now! . [root@bigboy tmp]# who root pts/0 Nov 6 14:46 bob pts/0 Nov 6 12:01 bunny pts/0 Nov 6 16:25 [root@bigboy tmp]# wall The system (192-168-1-242. PEs are important as you usually have to specify the size of your volume group not in gigabytes. Use the who command to see who’s logged in. therefore. Each logical volume can be individually formatted as if it were a regular Linux partition. The device /dev/hdf has a single partition named /dev/hdf1 into which /dev/hde5 will be merged. Another advantage of LVM is that this can all be done without disturbing other partitions on your hard disks. A logical volume is. The /home filesystem. like a virtual partition on your virtual disk drive. you’ll need to do some preparatory work before unmounting the filesystem: 1.my-web-site. which resides on /dev/hde5. but as a number of physical extents.my-web-site.org) (192-168-1-248. Take a look at the required steps.qxd 2/25/05 10:18 AM Page 443 Expanding Partitions with LVM 443 ☞ Volume Group: Any number of physical volumes (PVs) on different disk drives can be lumped together into a volume group (VG). send a message with the wall command informing them that the system is about to shutdown.my-web-site. has become too full. ☞ Logical Volumes: Volume groups must then be subdivided into logical volumes. but it allows you to create new virtual partitions with sizes you can change from groups of real disk partitions whose sizes you probably cannot change.30Harrison_ch27. Back up Your Data Use the tar command or some other method to backup your data in /home.

Log into the VGA console. DOS FDISK. Use it to modify both partitions /dev/hde5 and /dev/hdf1.05b# fdisk -l /dev/hde Disk /dev/hde: 4311 MB.. repeat them for /dev/hdf1.g. Here is an example using /dev/hde that shows your target partitions are of the incorrect type: sh-2. old versions of LILO) 2) booting and partitioning software from other OSs (e. and enter single-user mode: [root@bigboy tmp]# init 1 3. There is nothing wrong with that. 4311982080 bytes 16 heads.05b# umount /home Now we’re ready to start modifying the partitions.qxd 2/25/05 10:18 AM Page 444 444 Expanding Disk Capacity Chapter 27 Broadcast message from root (pts/0) (Sun Nov The system is shutting down now! [root@bigboy tmp]# 7 15:04:27 2004): 2. Determine the Partition Types You have to change each LVM partition used to be of type 8e (Linux LVM).05b# Start 1 4089 5714 6608 6608 End 4088 5713 6607 8355 7500 Blocks 2060320+ 819000 450576 880992 450040+ Id fd 83 83 5 83 System Linux raid autodetect Linux Linux Extended Linux Start fdisk You can change the partition type using fdisk with the disk name as its argument. 8355 cylinders Units = cylinders of 1008 * 512 = 516096 bytes Device Boot /dev/hde1 /dev/hde2 /dev/hde3 /dev/hde4 /dev/hde5 sh-2. sh-2. You can test this with the fdisk -l command..g. which is covered next. but this is larger than 1024.30Harrison_ch27. 63 sectors/track. and could in certain setups cause problems with: 1) software that runs at boot time (e. The fdisk examples that follow are for /dev/hde5. Unmount the filesystem: sh-2. OS/2 FDISK) Command (m for help): .05b# fdisk /dev/hde The number of cylinders for this disk is set to 8355.

63 sectors/track. 4311982080 bytes 16 heads. You can also use the L command to get a full listing of ID types in case you forget.30Harrison_ch27. The kernel still uses the old table.qxd 2/25/05 10:18 AM Page 445 Expanding Partitions with LVM 445 Set the ID Type to 8e You now need to set the partition types to the LVM value of 8e. 8355 cylinders Units = cylinders of 1008 * 512 = 516096 bytes Device Boot /dev/hde1 /dev/hde2 /dev/hde3 /dev/hde4 /dev/hde5 Start 1 4089 5714 6608 6608 End 4088 5713 6607 8355 7500 Blocks 2060320+ 819000 450576 880992 450040+ Id fd 83 83 5 8e System Linux raid autodetect Linux Linux Extended Linux LVM Command (m for help): Save the Partition Changes Use the w command to permanently save the changes to disk /dev/hde: Command (m for help): w The partition table has been altered! Calling ioctl() to re-read partition table. . WARNING: Re-reading the partition table failed with error 16: Device or resource busy. and then specify the partition number and type code. Partitions /dev/hde5 and /dev/hdf1 are the fifth and sixth partitions on disk /dev/hde. Command (m for help): t Partition number (1-6): 5 Hex code (type L to list codes): 8e Changed system type of partition 5 to 8e (Linux LVM) Command (m for help): t Partition number (1-6): 6 Hex code (type L to list codes): 8e Changed system type of partition 6 to 8e (Linux LVM) Command (m for help): Make Sure the Change Occurred Use the p command to get the new proposed partition table: Command (m for help): p Disk /dev/hde: 4311 MB. Modify their type using the t command.

05b# vgscan vgscan -. the volume group is called lvm-hde.05b# vgcreate lvm-hde /dev/hdf1 /dev/hde5 Volume group “lvm-hde” successfully created sh-2. Syncing disks. To do this..qxd 2/25/05 10:18 AM Page 446 446 Expanding Disk Capacity Chapter 27 The new table will be used at the next reboot.05b# Run VGscan The next step is to make Linux scan for any new LVM disk partitions and automatically create the LVM configuration files in the /etc directory. do it now! sh-2.physical volume “/dev/hde5” successfully created sh-2. This shouldn’t be grave as you are already in single user mode in which most of the system’s processes that would be accessing the partition have been shutdown. [root@bigboy updates]# The error above will occur if any of the other partitions on the disk is mounted. initialize the target partitions with the pvcreate command. This wipes out all the data on them in preparation for the next step.05b# Create a Volume Group for the PVs Use the vgcreate command to combine the two physical volumes into a single unit called a volume group.physical volume “/dev/hdf1” successfully created sh-2. The LVM software effectively tricks the operating system into thinking the volume group is a new hard disk.05b# pvcreate /dev/hdf1 pvcreate -.05b# .05b# pvcreate /dev/hde5 pvcreate -.) sh-2. If you haven’t backed up your data yet. sh-2. Define Each Physical Volume After modifying the partition tables of /dev/hde and /dev/hdf. use the vgscan command: sh-2.30Harrison_ch27. In the example..reading all physical volumes (this may take a while.

05b# Format the Logical Volume After the logical volume is created.Volume group --VG Name lvm-hde VG Access read/write VG Status available/resizable VG # 0 MAX LV 256 Cur LV 0 Open LV 0 MAX LV Size 255. logical volumes are divided into units called physical extents (PEs). You’ll have to know the number of available PEs before creating the logical volume.05b# lvcreate -l 212 lvm-hde -n lvm0 Logical volume “lvm0” created sh-2. This is done with the vgdisplay command using the new lvm-hde volume group as the argument: sh-2.05b# As you can see. In this case. Like hard disks.32 (09-Nov-2002) . 212 PEs are available as free. the vgcreate syntax uses the name of the volume group as the first argument followed by the partitions that it will be comprised of as all subsequent arguments.05b# vgdisplay lvm-hde --.99 GB Max PV 256 Cur PV 2 Act PV 2 VG Size 848 MB PE Size 4 MB Total PE 212 Alloc PE / Size 0 / 0 Free PE / Size 212 / 848 MB VG UUID W7bgLB-lAFW-wtKi-wZET-jDJF-8VYD-snUaSZ sh-2.30Harrison_ch27. use the -t switch to specify to the mkfs formatting program that you want a type ext3 partition: sh-2. You can now use all 212 of them to create a logical volume named lvm0 from volume group lvm-hde. which are divided into blocks of data. Create a Logical Volume from the Volume Group Now you’re ready to partition the volume group into logical volumes with the lvcreate command.05b# mkfs -t ext3 /dev/lvm-hde/lvm0 mke2fs 1. sh-2. you can format it as if it were a regular partition.qxd 2/25/05 10:18 AM Page 447 Expanding Partitions with LVM 447 Therefore.

qxd 2/25/05 10:18 AM Page 448 448 Expanding Disk Capacity Chapter 27 Filesystem label= OS type: Linux Block size=4096 (log=2) Fragment size=4096 (log=2) 108640 inodes. sh-2.00%) reserved for the super user First data block=0 7 block groups 32768 blocks per group. You. therefore. you lost the /home directory.05b# mkdir /home Update the /etc/fstab File The /etc/fstab file lists all the partitions that need to be automatically mounted when the system boots.05b# Create a Mount Point When you formatted the /dev/hde5 partition. don’t want the old partitions to be mounted again. whichever comes first. #/dev/hde5 #/dev/hdf1 /data1 /data2 ext3 ext3 defaults defaults 1 2 1 2 . This snippet configures the newly labeled partition to be mounted on the /home mount point: /dev/lvm-hde/lvm0 /home ext3 defaults 1 2 /lvm0 The /dev/hde5 and /dev/hdf1 partitions are replaced by the combined logical volume. Now you have to recreate /home on which you’ll later mount your new logical volume: sh-2. 217088 blocks 10854 blocks (5. 163840 Writing inode tables: done Creating journal (4096 blocks): done Writing superblocks and filesystem accounting information: done This filesystem will be automatically checked every 38 mounts or 180 days.30Harrison_ch27. 98304. 32768 fragments per group 15520 inodes per group Superblock backups stored on blocks: 32768. Use tune2fs -c or -i to override. Make sure that any reference to them in this file has either been commented with a # character at the beginning of each line or deleted entirely.

05b# Restore Your Data You can now restore your backed up data to /home. Expanding the existing disk capacity of your server may become an everyday occurrence and the tools described in this chapter should make the task easier. They will be around for a long time to come. Get Out of Single User Mode Return to your original run state by using either the init 3 or init 5 commands. denser memory chips were supposed to eliminate their need. It should just contain the lost+found directory. so does the data they are intended to store. dramatically increasing their speed and capacity too. Faster. It seems as if when drives get bigger. test the volume by listing its directory contents. .30Harrison_ch27.qxd 2/25/05 10:18 AM Page 449 Conclusion 449 Mount the Volume The mount -a command reads the /etc/fstab file and mounts all the devices that haven’t been mounted already. After mounting. sh-2. but hard disk technology has evolved.05b# ls /home lost+found sh-2.05b# mount -a sh-2. The exit command will make you return to your default runlevel. CONCLUSION The demise of the hard disk has been predicted for many years.

qxd 2/25/05 10:18 AM Page 450 .30Harrison_ch27.

and the setup is fairly simple. Use the who command to see which users are logged in. If you are certain that you’re the only user on the system. so wait until after hours to do this in a business environment. you may need to restrict the amount of disk space used on each partition by each user or group of users as your disk drives become filled with data. besides yourself. The procedure is quick: 1. you might be able to skip this step. send a message stating that the system is about to shutdown with the wall command: 451 /home . Enter Single-User Mode As you’ll need to remount the /home filesystem.qxd 2/25/05 10:19 AM Page 451 C H A P T E R 28 Managing Disk Usage with Quotas IN THIS CHAPTER ☞ Setting Up Quotas ☞ Other Quota Topics ☞ Conclusion Eventually. If there are any. The disk quota feature of Red Hat/Fedora Linux enables you to do this. SETTING UP QUOTAS For example. your family Linux server is running out of space in the filesystem because of an abundance of MP3 downloads. you make sure that no other users or processes are using it by first entering single-user mode from the console. Entering single-user mode automatically logs off all users and stops cron jobs.31Harrison_ch28.

org) 6 12:01 (192-168-1-248. Linux needs to reread the file to get its instructions for /home. The old fstab looked like: LABEL=/home /home ext3 defaults 1 2 but your new fstab should be: LABEL=/home /home ext3 defaults. You can do this using the mount command with the -o remount qualifier.05b# mount -o remount /home Get Out of Single-User Mode Return to your original run state by using either the init 3 or init 5 commands.org) system is shutting down now! 7 15:04:27 2004): Broadcast message from root (pts/0) (Sun Nov The system is shutting down now! [root@bigboy tmp]# 2. and enter single user mode: [root@bigboy tmp]# init 1 Simple.org) 6 16:25 (192-168-1-250.31Harrison_ch28.usrquota 1 2 Remount the Filesystem Editing the /etc/fstab file isn’t enough. You have to alert Linux that quotas are enabled on the filesystem by editing the /etc/fstab file and modifying the options for the /home directory. the usrquota option is mentioned in the fstab man pages. . Log into the VGA console. Continue to the next step once the system is back to its normal state. You’ll need to add the usrquota option. sh-2.qxd 2/25/05 10:19 AM Page 452 452 Managing Disk Usage with Quotas Chapter 28 [root@bigboy tmp]# who root pts/0 Nov bob pts/0 Nov bunny pts/0 Nov [root@bigboy tmp]# wall The 6 14:46 (192-168-1-242. eh? Edit Your /etc/fstab File The /etc/fstab file lists all the partitions that need to be auto-mounted when the system boots. You can also use the exit command to return to your default runlevel.my-web-site.my-web-site.my-web-site. In case you forget the name.

just enable per-user quotas for the /home filesystem: [root@bigboy tmp]# touch /home/aquota.user was probably truncated. The edquota command enables you to selectively edit a portion of the aquota. an aquota. This table will be automatically and transparently updated each time a file is modified. or both. quotacheck: Scanning /dev/hda3 [/home] done quotacheck: Checked 185 directories and 926 files [root@bigboy tmp]# Edit the User’s Quota Information Now you need to edit the user’s quota information. Linux compares the values in this table with the quota limitations that the systems administrator has placed in the aquota files and uses this information to determine whether the user has rights to increased disk usage. Be prepared: You’ll get an error the first time you enter the command. Disk quotas for user mp3user (uid 503): Filesystem blocks soft hard /dev/hda3 24 0 0 inodes 7 soft 0 hard 0 .Quotafile /home/aquota.user file on a per-user basis: [root@bigboy tmp]# edquota -u mp3user The command invokes the vi editor.user file (defines quotas by user).user [root@bigboy tmp]# chmod 600 /home/aquota. You have to generate a quota table. Can’t save quota settings..31Harrison_ch28. because Linux will realize that the aquota file wasn’t created using one of the quota commands: [root@bigboy tmp]# quotacheck -vagum quotacheck: WARNING .group file (defines quotas by group). In this case.user [root@bigboy tmp]# Initialize the Quota Table Editing the /etc/fstab file and remounting the file system only alerted Linux to the fact that the filesystem has quota capabilities. The man page for quota lists them at the bottom. that lists all the current allocations for each user on the file system. separate from the aquota files.. You initialize the table with the quotacheck command.qxd 2/25/05 10:19 AM Page 453 Setting Up Quotas 453 Create the Partition Quota Configuration Files The uppermost directory of the filesystem needs to have an aquota.

days. Like the The grace period is a time limit before the soft limit is enforced for a quota-enabled file system. When the grace period expires. depending on the configuration. ☞ Hard limit: The maximum blocks/inodes a quota user may have on a partition when a grace period is set. it invokes the vi editor. You’ll need to make Linux scan its hard disks periodically to check for exceeded quotas. the user is warned only that their soft limit has been exceeded. ☞ Soft limit: The maximum blocks/inodes a quota user may have on a partition. and months. When set to zero. then Linux prevents the creation of new files or the expansion of existing files to use more disk space. limits are disabled. You can use time units of seconds. but they can never exceed their hard limit. Here user mp3user is limited to a maximum of 5MB of data storage on /dev/hda3 (/home): Disk quotas for user mp3user (uid 503): Filesystem blocks soft hard /dev/hda3 24 5000 0 inodes 7 soft 0 hard 0 Testing Linux checks the total amount of disk space a user uses each time a file is accessed and compares it against the values in the quota file. OTHER QUOTA TOPICS Creating disk quotas frequently isn’t enough.qxd 2/25/05 10:19 AM Page 454 454 Managing Disk Usage with Quotas Chapter 28 From here. in 1K blocks. minutes. With grace periods. hours. This is what you’ll see with the command edquota -t: . The for each filesystem. ☞ Inodes: The number of files the user is currently using. The role of a soft limit changes if you use grace periods. that the user is currently using. Editing Grace Periods edquota -t command sets the grace period edquota -u command. This section describes the most common quota management activities. Users may exceed a soft limit.31Harrison_ch28. the user is barred from using additional disk space or files. If the values are exceeded. weeks. you can edit a number of fields: ☞ Blocks: The amount of space. You also have to manage the process by reviewing the quota needs of each user and adjusting them according to the policies of your company.

but also reduces the likelihood of having fragmented disk structures that slow disk access due to the presence of too many files. Therefore. Getting Quota Reports The repquota command lists quota usage limits of all users on the system: [root@bigboy tmp]# repquota /home *** Report for user quotas on device /dev/hda3 Block grace time: 7days. Of course.. It helps to not only limit the use of physical disk resources that could potentially be put to better use. “Expanding Disk Capacity. Chapter 27. hours. 7days is correct. . . minutes.31Harrison_ch28.qxd 2/25/05 10:19 AM Page 455 Other Quota Topics 455 [root@bigboy tmp]# edquota -t Grace period before enforcing soft limits for users: Time units may be: days.. mp3user -24 0 0 7 0 0 [root@bigboy tmp]# -g CONCLUSION Disk quotas are good to put in place for specific users whose disk usage activities need to be curtailed... users must be made aware of the challenges you face as a result of excessive disk usage.” covers a variety of approaches to increasing your storage capacity. User education must play a major role in your strategy too.. Editing Group Quotas Editing quotas on a per-group basis can be done similarly with the edquota command. there comes a time when you are faced with no option but to expand the number of disks in your system. .. or seconds Filesystem Block grace period Inode grace period /dev/hda3 7days 7days Note There should be no spaces between the number and the unit of time measure. Inode grace time: 7days Block limits File limits User used soft hard grace used soft hard grace ---------------------------------------------------------------------root -52696 0 0 1015 0 0 . but 7 days is wrong.

qxd 2/25/05 10:19 AM Page 456 .31Harrison_ch28.

New disks are added by attaching. Basic configuration is fairly simple.qxd 2/25/05 10:19 AM Page 457 C H A P T E R 29 Remote Disk Access With NFS IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ NFS Operation Overview Installing nfs Scenario Configuring NFS on the Server Configuring NFS on the Client Activating Modifications to the /etc/exports File The NFS automounter Troubleshooting NFS Other NFS Considerations Conclusion Samba is usually the solution of choice when you want to share disk space between Linux and Windows machines. the directories of their filesystems to a directory of an already existing filesystem. The Network File System protocol (NFS) is used when disks need to be shared between Linux servers. and this chapter will explain all the essential steps. or mounting. The systems administrator on the NFS server has to define the 457 . NFS was developed to allow a computer system to access directories on remote computers by mounting them on a local filesystem as if they were a local disk.32Harrison_ch29. This in effect makes the new hard disk transparently appear to be a subdirectory of the filesystem to which it is attached. NFS OPERATION OVERVIEW Linux data storage disks contain files stored in filesystems with a standardized directory structure.

Key NFS Concepts Data access over a network always introduces a variety of challenges. ☞ Do not export a subdirectory of a directory that has already been exported—the exception being when the subdirectory is on a different physical device. Stateless Operation Programs that read and write to files on a local filesystem rely on the operating system to track their access location within the file with a pointer. or obscured. the NFS server doesn’t maintain the file pointer . This is done in such a way that files on the remote NFS server appear to the user to be no different than those on a local disk. As NFS is a network-based file system. do not export the parent of a subdirectory unless it is on a separate device. when a server fails. it was decided that the NFS client daemon would act as a failsafe intermediary between regular programs running on the NFS client and the NFS server. VFS also translates these requests to match the actual filesystem format on the NFS server’s disks. ☞ Only export local filesystems. Likewise. file accesses timeout and the file pointers are reset to zero. Here are some key NFS background concepts that will help in your overall understanding. General NFS Rules You should follow some general rules when configuring NFS: ☞ Only export directories beneath the / directory. This allows the NFS server to be not only a completely different operating system.qxd 2/25/05 10:19 AM Page 458 458 Remote Disk Access With NFS Chapter 29 directories that need to be activated. With NFS. for access by the NFS clients. Normally. VFS The virtual filesystem (VFS) interface is the mechanism used by NFS to transparently and automatically redirect all access to NFS-mounted files to the remote server. as in the case of NFS. Keep in mind that when you mount any filesystem on a directory. or exported. When the filesystem is unmounted. and administrators on the clients need to define both the NFS server and the subset of its exported directories to use. then the original files in the directory reappear unchanged. especially if the operation is intended to be transparent to the user.32Harrison_ch29. and networks can be unreliable. the original contents of the directory are ignored. but also use different naming schemes for files with different file attribute types. in favor of the files in the mounted filesystem.

This is also known as a read ahead cache. Data that’s to be written to the NFS server is cached with the data being written to the server when the cache becomes full. whether in the background or foreground. the mount command will hang for a week until it returns online.. then an error will occur. relative symbolic links are interpreted relative to the client’s filesystem. NFS Background Mounting NFS clients use the remote procedure call (RPC) suite of network application helper programs to mount remote filesystems. The NFS server caches information too. This means that if an NFS server suddenly fails. mounting a filesystem on a symbolic link actually mounts the filesystem on the target of the symbolic link. If there is a link to the .32Harrison_ch29. It is possible to use a bg option to spawn the retries off as a subprocess so that the main mount command can continue to process other requests./data2 directory on the NFS server and a directory corresponding to . Also.000 minutes. The difficulty here is that if the NFS server is unavailable. The default is 10. With soft mounts. If the mount cannot occur during the default RPC timeout period. then the client retries the mount process until the NFS number of retries has been exceeded.qxd 2/25/05 10:19 AM Page 459 NFS Operation Overview 459 information. is called a hard mount. the NFS client does. then the NFS client won’t be able to access it. which is approximately a week. the NFS client can precisely restart the file access once more after patiently waiting until the server returns online. repeated RPC failures . Consider an example where the /data1 directory on the server is mounted on the /data1 directory. Hard and Soft Mounts The process of continuous retrying. NFS and Symbolic Links You have to be careful with the use of symbolic links on exported NFS directories. You’ll have to be careful not to obscure the contents of this original directory in the process. Caching NFS clients typically request more data than they need and cache the results in memory locally so that further sequential access of the data can be done locally versus over the network. NFS attempts to guarantee the consistency of your data with these constant retries. Plan carefully before doing this. Caching therefore helps to reduce the amount of network traffic while simultaneously improving the speed of some types of data access. such as the directory information for the most recently accessed files and a read ahead cache for recently read files. Unlike absolute links./data2 doesn’t exist on the NFS client.. If an absolute link points to a directory on the NFS server that hasn’t been exported.

whereas before it was optional ☞ Support for the bundling of requests from each client provides more efficient processing by the NFS server It is important to match the versions of NFS running on clients and servers to help ensure the necessary compatibility to get NFS to work predictably. Important NFS Daemons NFS isn’t a single program. but a suite of interrelated programs that work together to get the job done: ☞ The primary daemon upon which all the others rely. The disadvantage is that the use of the soft option implies that you are using an unreliable NFS server. NFS Versions Three versions of NFS are available currently: versions 2. This is then used to negotiate a range of TCP ports. 3. portmap listens to TCP port 111 on which an initial connection is made. whether it fails or not. but with the additions that: ☞ File locking and mounting are integrated in the NFS daemon and operate on a single. usually above portmap: .32Harrison_ch29. This chapter focuses on version 2. if this is the case it is best not to place critical data that needs to be updated regularly or executable programs in such a location. portmap manages connections for applications that use the RPC specification.1 bytes ☞ Supports the NFS server data updates as being successful when the data is written to the server’s cache ☞ Negotiates the data limit per read or write request between the client and server to a mutually decided optimal value Version 4 maintains many of version 3’s features. Version 1 was a prototype. and 4. By default. The advantage is that the operation completes quickly. making network security easier ☞ File locking is mandatory. which: ☞ Supports files up to 4GB long ☞ Requires an NFS server to successfully write data to its disks before the write request is considered successful ☞ Has a limit of 8KB per read or write request The main differences with version 3 are that it: ☞ Supports extremely large file sizes of up to 264 . well known TCP port.qxd 2/25/05 10:19 AM Page 460 460 Remote Disk Access With NFS Chapter 29 cause the NFS operation to fail not hang and data consistency is therefore not guaranteed.

1. You can determine whether you have nfs installed using the RPM command in conjunction with the grep command to search for all installed nfs packages: [root@bigboy tmp]# rpm -qa | grep nfs redhat-config-nfs-1. see Chapter 6.i386. You also need to have the RPC portmap package installed. Now take a look at how to configure these daemons to create functional NFS client/server peering. If nfs and portmap are not installed.1. You need to run portmap on both the NFS server and client. ☞ nfs: Starts the RPC processes needed to serve shared NFS file systems.rpm. you can determine all the portmap applications installed: [root@bigboy tmp]# rpm -q portmap portmap-4. INSTALLING NFS Red Hat Linux installs nfs by default. “Installing RPM Software. they can be added fairly easily once you find the nfs-utils and portmap RPMs.0-57 [root@bigboy tmp]# A blank list means that you’ll have to install the required packages.”) Remember that RPM filenames usually start with the software’s name and a version number.32Harrison_ch29. The nfslock daemon needs to be run on both the NFS server and client. . When you use rpm in conjunction with grep. to be used for subsequent data transfers. and the rpm command can tell you whether it’s on your system already. ☞ netfs: Allows RPC processes run on NFS clients to mount NFS filesystems on the server. and also by default nfs is activated when the system boots.9 [root@bigboy tmp]# A blank list means that you’ll have to install the required packages. The nfslock daemon needs to be run on the NFS client only.0.3-1 nfs-utils-1.qxd 2/25/05 10:19 AM Page 461 Installing nfs 461 port 1024. The nfs daemon needs to be run on the NFS server only.31.1-3. (If you need a refresher. ☞ nfslock: Used to allow NFS clients to lock files on the server via RPC processes. as in nfs-utils-1.

32Harrison_ch29.168. The first column lists the directories you want to make available to the network.203/32(rw.255 ☞ Read/write access to the /data/test directory from servers in the my-web-site.org DNS domain ☞ Read/write access to the /data/database directory from a single server. even after hours. as well as a correctly configured /etc/exports file. 192.qxd 2/25/05 10:19 AM Page 462 462 Remote Disk Access With NFS Chapter 29 SCENARIO A small office has an old Linux server that is running out of disk space. The /etc/exports File The /etc/exports file is the main NFS configuration file.168.0 to 192.1. but first make sure that NFS is running correctly.168. because the server is accessed by overseas programmers and clients at night and local ones by day.1.1.org(rw. and the second part lists NFS options in brackets.sync) 192.my-web-site.1.sync) 192. Another Linux server on the network has additional disk capacity in its /data partition and the office would like to expand into it as an interim expansion NFS server. and it consists of two columns.168.0 /24 network. which is all addresses from 192. nfs. The first part lists the networks or DNS domains that can get access to the directory.1.168. The second column has two parts.168.1. The server needs portmap. use the sync option to ensure that file data cached in memory is automatically written to the disk after the completion of any disk data copying operation. and nfslock operational. The office cannot tolerate any down time.203 In all cases. Here’s how to do it. CONFIGURING NFS ON THE SERVER Both the NFS server and NFS client have to have parts of the NFS package installed and running. Budgets are tight and the company needs a quick solution until it can get a purchase order approved for a hardware upgrade.sync) *. For the scenario you need: ☞ Read-only access to the /data/files directory to all networks ☞ Read/write access to the /home directory from all servers on the 192. .sync) After configuring your /etc/exports file.0/24(rw. #/etc/exports /data/files /home /data/test /data/database *(ro. you need to activate the settings.

you can also stop and restart the processes with the stop and restart options: [root@bigboy tmp]# service portmap start [root@bigboy tmp]# service nfs start [root@bigboy tmp]# service nfslock start 3. or mount point. create a directory on which to mount the NFS server’s directories that you exported via the /etc/exports file. Use the chkconfig command to configure the required nfs and RPC portmap daemons to start at boot. nfs. and finally to mount the NFS server’s directory on your local directory. and nlockmgr: [root@bigboy tmp]# rpcinfo -p localhost program vers proto port 100000 2 tcp 111 portmapper 100000 2 udp 111 portmapper 100003 2 udp 2049 nfs 100003 3 udp 2049 nfs 100021 1 udp 1024 nlockmgr 100021 3 udp 1024 nlockmgr 100021 4 udp 1024 nlockmgr 100005 1 udp 1042 mountd 100005 1 tcp 2342 mountd 100005 2 udp 1042 mountd 100005 2 tcp 2342 mountd 100005 3 udp 1042 mountd 100005 3 tcp 2342 mountd [root@bigboy tmp]# CONFIGURING NFS ON THE CLIENT NFS configuration on the client requires you to start the NFS application. The examples use the start option.d directory to start the nfs and RPC portmap daemons. but when needed. You should get a listing of running RPC programs that must include mountd. You also should activate NFS file locking to reduce the risk of corrupted data. Use the init scripts in the /etc/init.qxd 2/25/05 10:19 AM Page 463 Configuring NFS on the Client 463 Starting NFS on the Server Configuring an NFS server is straightforward: 1.32Harrison_ch29. Here’s how to do it all. . [root@bigboy tmp]# chkconfig --level 35 nfs on [root@bigboy tmp]# chkconfig --level 35 nfslock on [root@bigboy tmp]# chkconfig --level 35 portmap on 2. Test whether NFS is running correctly with the rpcinfo command. portmapper.

in-addr.168. Test whether NFS is running correctly with the rpcinfo command. [root@bigboy tmp]# host 192.192. As on the server.102 201. In other words.org. the examples use the start option.qxd 2/25/05 10:19 AM Page 464 464 Remote Disk Access With NFS Chapter 29 Starting NFS on the Client Three more steps easily configure NFS on the client: 1.168. The listing of running RPC programs you get must include status. Activate nfslock to lock the files and reduce the risk of corrupted data: [root@smallfry tmp]# chkconfig --level 35 netfs on [root@smallfry tmp]# chkconfig --level 35 nfslock on [root@smallfry tmp]# chkconfig --level 35 portmap on 2. a DNS lookup on the NFS server for the IP address of the NFS client must return a server name that will map back to the original IP address when a DNS lookup is done on that same server name. and nlockmgr: [root@smallfry root]# rpcinfo -p program vers proto port 100000 2 tcp 111 portmapper 100000 2 udp 111 portmapper 100024 1 udp 32768 status 100024 1 tcp 32768 status 100021 1 udp 32769 nlockmgr 100021 3 udp 32769 nlockmgr 100021 4 udp 32769 nlockmgr 100021 1 tcp 32769 nlockmgr 100021 3 tcp 32769 nlockmgr 100021 4 tcp 32769 nlockmgr 391002 2 tcp 32770 sgi_fam [root@smallfry root]# NFS and DNS The NFS client must have a matching pair of forward and reverse DNS entries on the DNS server used by the NFS server.d directory to start the nfs and RPC portmap daemons. . Use the init scripts in the /etc/init. but you can also stop and restart the processes with the stop and restart options: [root@smallfry tmp]# service portmap start [root@smallfry tmp]# service netfs start [root@smallfry tmp]# service nfslock start 3. Use the chkconfig command to configure the required nfs and RPC portmap daemons to start at boot.32Harrison_ch29.arpa domain name pointer 192-168-1-102.1. portmapper.my-website.1.

100) as an NFS-type filesystem using the local /mnt/nfs mount point directory: #/etc/fstab #Directory FSCK 192.168. you need to edit the /etc/fstab file if you need the NFS directory to be made permanently available to users on the NFS. See the NFS man pages for more details. Use soft mounting. The /etc/fstab File The /etc/fstab file lists all the partitions that need to be auto-mounted when the system boots.168.16801. mount the /data/files directory on server Bigboy (IP address 192. Table 29.168.1.org has address 192.my-web-site. This requires an entry in the /etc/fstab file in addition to the creation of the mount point directory.org 192-168-1-102.1. For the example.mountd: Fake hostname smallfry. users want their NFS directories to be permanently mounted.102 [root@bigboy tmp]# This is a security precaution added into the nfs package that lessens the likelihood of unauthorized servers from gaining access to files on the NFS server.my-web-site. Use hard mounting.my-website.1 outlines these and other useful NFS mounting options you may want to use.nfsvers=2 This example used the soft and nfsvers options.32Harrison_ch29.1 .qxd 2/25/05 10:19 AM Page 465 Configuring NFS on the Client 465 [root@bigboy tmp]# host 192-168-1-102.1 Option bg fg soft hard Possible NFS Mount Options Description Retry mounting in the background if mounting initially fails. continues . Therefore. Mount in the foreground. Failure to correctly register your server IPs in DNS can result in “fake hostname” errors: Nov 7 19:14:40 bigboy rpc.1.100:/data/files 0 Mount Point /mnt/nfs Type nfs Options Dump 0 soft.org for 192.forward lookup doesn’t exist Making NFS Mounting Permanent In most cases. Table 29.

the most popular of which is version 2. this changes after the mounting is completed: [root@smallfry tmp]# mkdir /mnt/nfs [root@smallfry tmp]# ls /mnt/nfs [root@smallfry tmp]# mount -a [root@smallfry tmp]# ls /mnt/nfs ISO ISO-RedHat kickstart RedHat [root@smallfry tmp]# mount -a Each time your system boots. on which to mount the remote NFS directory and then use the mount -a command to activate the mount.1. For NFS version 2. Note There are multiple versions of NFS.32Harrison_ch29. Permanently Mounting the NFS Directory You’ll now create a mount point directory. The default is dependent on the kernel.100’ failed: server is down. Failure to do so may result in an error message: [root@probe-001 tmp]# mount -a mount to NFS server ‘192. wsize=n nfsvers=n tcp intr The steps to mount the directory are fairly simple. allow for the process to be aborted using the usual methods such as CTRL-C and the kill command. Notice how before mounting there were no files visible in the /mnt/nfs directory. Attempt to mount the filesystem using TCP packets. The amount of data NFS will attempt to access per write operation.168. To be safe. thereby making this a permanent NFS mount. The version of NFS the mount command should attempt to use. it reads the /etc/fstab file and executes the command. it is best to force the NFS server to export directories as version 2 using the nfsvers=2 option in the /etc/fstab file as shown in the example. If the filesystem is hard mounted and the mount times out. as you’ll see. the default is UDP. Newer NFS servers may also be able to handle NFS version 4. set it to 8192 to assure maximum throughput. The default is dependent on the kernel.1 Option rsize=n continued Description The amount of data NFS will attempt to access per read operation. which most NFS clients use.qxd 2/25/05 10:19 AM Page 466 466 Remote Disk Access With NFS Chapter 29 Table 29. set it to 8192 to assure maximum throughput. [root@probe-001 tmp]# . For NFS version 2. /mnt/nfs.

1. Here are some methods you can use to update and activate the file with the least amount of inconvenience to others.qxd 2/25/05 10:19 AM Page 467 Activating Modifications to the /etc/exports File 467 Manually Mounting NFS File Systems If you don’t want a permanent NFS mount.” In this case. see the section “The NFS automounter. Notice how before mounting there were no files visible in the /mnt/nfs directory. ACTIVATING MODIFICATIONS TO THE /ETC/EXPORTS FILE You can force your system to re-read the /etc/exports file by restarting NFS.1. this may cause disruptions when an exported directory suddenly disappears without prior notification to users. New Exports File When no directories have yet been exported to NFS. This is a manual process. for an automated process.100. you’re mounting the /data/files directory as an NFS-type filesystem on the /mnt/nfs mount point. this changes after the mounting is complete: [root@smallfry tmp]# mkdir /mnt/nfs [root@smallfry tmp]# ls /mnt/nfs [root@smallfry tmp]# mount -t nfs 192. you can use the export only the new entries: [root@bigboy tmp]# exportfs -r exportfs -r command to .32Harrison_ch29. then you can use the mount command without the /etc/fstab entry to gain access only when necessary. The NFS server is Bigboy whose IP address is 192.168. In a nonproduction environment.100:/data/files /mnt/nfs [root@smallfry tmp]# ls /mnt/nfs ISO ISO-RedHat kickstart RedHat [root@smallfry tmp]# Congratulations! You’ve made your first steps towards being an NFS administrator.168. use the command: [root@bigboy tmp]# exportfs -a exportfs -a Adding a Shared Directory To An Existing Exports File When adding a shared directory.

the mount is permanent. tying up system resources even when the NFS server isn’t being accessed. Unexport the mount point directory on the NFS client using the umount command.32Harrison_ch29. automounter intercepts it and mounts the remote directory on the mount point defined in the map file. However. automounter continues to report to the operating system kernel that the mount is still active.qxd 2/25/05 10:19 AM Page 468 468 Remote Disk Access With NFS Chapter 29 Deleting. In this case. instead using an NFS-specific map file that can be distributed to multiple clients. Moving. Also. after the timeout period the remote directory is unmounted and the mount point is deleted. NFS uses an automounter feature that overcomes these shortcomings by allowing you to bypass the /etc/fstab file for NFS mounts. 2. The steps are: 1. the /etc/fstab file is unique per Linux server and has to be individually edited on each. For example. Comment out the corresponding entry in the NFS server’s file and reload the modified file: [root@bigboy tmp]# exportfs -ua [root@bigboy tmp]# exportfs -a /etc/exports You have now completed a seamless removal of the exported directory with much less chance of having critical errors. automounter Map Files The master map file of automounter has a simple format that defines the name of the mount point directory in the first column and the subsidiary map file that controls its mounting in the second. You can add mounting options to a third column. When the kernel makes an NFS file request. . In addition. The mount point directory is dynamically created by the automounter when needed. you can use the file to specify the expected duration of the NFS mount. becomes more difficult. THE NFS AUTOMOUNTER The permanent mounting of filesystems has its disadvantages. you’re unmounting the /mnt/nfs mount point: [root@smallfry tmp]# umount /mnt/nfs Note You may also need to edit the /etc/fstab file of any entries related to the mount point if you want to make the change permanent even after rebooting. therefore. after which time it is unmounted automatically. or Modifying a Share Removing an exported directory from the /etc/exports file requires work on both the NFS client and server. NFS client management.

You have to have one indirect map defined to avoid startup errors. The Structure of Direct and Indirect Map Files The format of these map files is similar to that of the /etc/auto. then automounter checks the subsidiary map file for a subdirectory mount point key. . If it finds one.master # /home /etc/auto. an early release of Fedora Core 3 doesn’t appear to work correctly with automounter.direct Direct Maps Direct maps are used to define NFS filesystems that are mounted on different servers or that don’t all start with the same prefix. A good example would be all the users’ directories under /home. Irregular entries that don’t match /home are placed in the /etc/auto.qxd 2/25/05 10:19 AM Page 469 The NFS automounter 469 In the example. and this value will act as a default for all the entries in the /etc/auto.direct file.master file. the /home directory needs to be automounted on an NFS server and the configuration information is defined in the /etc/auto. No errors occur in the logs either. If one does. then automounter mounts the files for the system. it refers to the keys in the /etc/auto. Indirect Maps Indirect maps define directories that can be mounted under the same mount point.32Harrison_ch29. Finally. It is also the name of the mount point under the directory listed in the /etc/auto.master file to see whether any fall within the realm of the automounter’s responsibility. Note Based on preliminary testing. except that columns two and three have been switched.home --timeout=300 //etc/auto.master file. the mount will only last for five minutes (300 seconds). When the NFS client accesses a file. Column one lists all the directory keys that will activate the automounter feature. # # File: /etc/auto.home file. The second column provides all the NFS options to be used. the maps don’t appear to be activated. and the third column lists the NFS servers and the filesystems that map to the keys.home file. and after doing so.

these directories are actually mount points for directories on servers Bigboy. if you don’t then you’ll get an error like this in your /var/log/messages file: Nov 7 19:24:12 smallfry automount[31801]: bad map format: found indirect.master file redirected all references to the /home directory to the /etc/auto.master file was specifically created to handle all references to one of a kind directory prefixes. and Waitabit. Ochorios. This means all direct mount points in the same directory are mounted simultaneously even if only one of them is being accessed.home bigboy:/home/peter ochorios:/home/bob waitabit:/home/bunny Direct Map File Example The second entry in the /etc/auto.home file. and bunny.master file must all begin with /-.32Harrison_ch29. This second file has entries for peter. There are tricks you can use to avoid this. In the example the /data/sales and /sql/database are the mount points for directories on servers Bigboy and Waitabit. and you can use absolute path names with direct map files only. bob.soft bigboy:/disk1/data/sales waitabit:/var/mysql/database Note The automounter treats direct mounts as if they were files in a directory. expected direct exiting .qxd 2/25/05 10:19 AM Page 470 470 Remote Disk Access With NFS Chapter 29 Indirect Map File Example In the previous example. # # File: /etc/auto. This can cause excessive mounting activity that can slow response times. not as individual directories. Note Direct map entries in the /etc/auto. perhaps the simplest is just to place direct mount points in different directories. # # File: # peter bob bunny /etc/auto.direct # /data/sales -rw /sql/database -ro. the /etc/auto.

# # File: /etc/auto. then the ampersand is interpreted to mean peter. so the ampersand wildcard is interpreted to mean peter too. called autofs. then bigboy:/home/bob would be mounted. # # File: /etc/auto.32Harrison_ch29.home # peter bigboy:/home/& Using the Asterisk Wildcard In the example below. and bigboy:/home/peter is mounted.home file.qxd 2/25/05 10:19 AM Page 471 The NFS automounter 471 Wildcards in Map Files You can use two types of wildcards in a map file. and the ampersand (&). the key is peter. which means all. meaning that automounter will attempt to mount any attempt to enter the /home directory. The asterisk (*). If the access was for /home/peter. If access was for /home/bob. This means you’ll be mounting the bigboy:/home/peter directory. Activate NFS file locking to reduce the risk of corrupted data.home # * bigboy:/home/& Starting automounter Fedora Linux installs the automounter RPM.d directory to start the automounter daemons. by default. which instructs automounter to substitute the value of the key for the & character. Using the Ampersand Wildcard In the example below. [root@bigboy tmp]# chkconfig autofs on 2. but you can also stop and restart the process with the stop and restart options. Here are some quick steps to get automounter started: 1. Use the chkconfig command to configure the automounter daemons to start at boot. The example uses the start option. the key is *. Use the init scripts in the /etc/init. [root@bigboy tmp]# service autofs start . But what’s the value of the ampersand? It is actually assigned the value of the key that triggered the access to the /etc/auto.

The most common problems are caused by forgetting to start NFS. to edit the /etc/fstab file. Typographical error in the name of the server. Another common cause of failure is the iptables firewall daemon running on either the server or client without the administrator realizing it. but you’ll also have to make sure basic connectivity has been taken care of as outlined in Chapter 4. As always. Table 29. and accounting control.qxd 2/25/05 10:19 AM Page 472 472 Remote Disk Access With NFS Chapter 29 3. When the client and server are on different networks. Typographical error in the name of the file or directory. or to export the /etc/exports file. TROUBLESHOOTING NFS A basic NFS configuration usually works without problems when the client and server are on the same network. they don’t exist. “Simple Network Troubleshooting. “Centralized Logins Using NIS. Option Too many levels of remote in path Permission denied No such host No such file or directory . Additional centralization is also achieved by using NIS for login authentication. the command will return the process ID of the automount daemon. access. Chapter 30. no troubleshooting plan would be complete without frequent reference to the /var/log/messages file when searching for additional clues. Table 29.32Harrison_ch29.2 Some Common NFS Error Messages Description Attempting to mount a filesystem that has already been mounted. This could be the client’s root user who has unprivileged status on the server due to the root_squash option. Could also be because the user on the client doesn’t exist on the server. automounter Examples Now that you understand NFS automounter.2 shows some common NFS errors you’ll encounter. User is denied access. managing the startup of automounter is very similar to that of other Linux applications and should be easy to remember. [root@bigboy tmp]# pgrep automount 32261 [root@bigboy tmp]# As you can see. If it is. Use the pgrep command to determine whether automounter is running.” Sometimes a firewall being present on the path between the client and server can cause difficulties.” contains a full scenario in which a school computer laboratory uses automounter to centrally house all the home directories of its students. these checks still apply. you may benefit from an example.

Forward and reverse DNS entries don’t exist for the NFS client.qxd 2/25/05 10:19 AM Page 473 Troubleshooting NFS 473 Table 29.100:/home/nfsuser 1032056 346552 633068 36% /home/nfsuser [nfsuser@smallfry nfsuser]$ The nfsstat Command The nfsstat command provides useful error statistics. [root@bigboy tmp]# nfsstat -s Server rpc stats: calls badcalls badauth 1547 0 0 Server nfs v2: null getattr setattr badclnt 0 root xdrcall 0 lookup readlink . Option NFS server is not responding Stale file handle Fake hostname The showmount Command When run on the server.1.3.32Harrison_ch29. while the -c option provides them for clients. In many cases. in this case one client has an IP address of 192.102: [root@bigboy tmp]# showmount -a All mount points on bigboy: *:/home 192. Threshold guidelines are provided in Table 29. It also shows a list of NFS clients accessing the server. the root_squash mount option will prevent the root user from doing this. so it’s best to try it as an unprivileged user.168. Run it on the NFS client to verify that NFS mounting has occurred.102:* [root@bigboy tmp]# The df Command The df command lists the disk usage of a mounted filesystem.2 Some Common NFS Error Messages Description The server could be overloaded or down.168. The -s option provides NFS server stats.1. [nfsuser@smallfry nfsuser]$ df -F nfs Filesystem 1K-blocks Used Available Use% Mounted on 192. the showmount -a command lists all the currently exported directories.1. A file that was previously accessed by the client was deleted on the server before the client closed it.168.

Usually caused by the NFS noac mount option. Percentage of requests for service that the client had to retransmit to the servers. Its security is relatively weak. Try to replace them with a directory and mount the filesystem directly on this new mount point.3 Value Readlink Error Thresholds for the nfsstat Command Threshold > 10% Description Excessive numbers of symbolic links slowing performance.32Harrison_ch29. Bad RPC requests. which prevents file attribute caching. Check the noac and wsize mount options. . An incorrect configuration can cause it to be unresponsive. the root user attempting to access data governed by the root_squash directive. File attributes. Could be due to slow NFS servers or poor network conditions. This value tracks the percentage of file attribute reads that are not from cache refresh requests. Getattr > 50% badcalls >0 retrans > 5% Writes Read > 10% > 10% OTHER NFS CONSIDERATIONS NFS can be temperamental.qxd 2/25/05 10:19 AM Page 474 474 Remote Disk Access With NFS Chapter 29 244 read 0 link 0 100% 0 0% 0 wrcache write 0% 0 0% 0 symlink mkdir 0% 0 0% 0 0% 0 create 0% 0 rmdir 0% 0 0% 0 0% 0 remove rename 0% 0 0% 0 readdir fsstat 0% 0 0% 0 0% 0% 0% Server nfs v3: null getattr setattr 251 19% 332 25% 0 0% read write create 39 2% 14 1% 1 0% remove rmdir rename 0 0% 0 0% 0 0% fsstat fsinfo pathconf 1 0% 34 2% 0 0% [root@bigboy tmp]# lookup access readlink 265 20% 320 24% 0 0% mkdir symlink mknod 1 0% 0 0% 0 0% link readdir readdirplus 0 0% 0 0% 31 2% commit 14 1% Table 29. Writes are slow due to poor caching values. Often these issues can be resolved with some basic guidelines outlined in this section. like file data. or having a user in too many groups. Check the noac and rsize mount options. is cached in NFS. and you have to be aware of the file permissions on both the NFS client and server to get it to work correctly. Reads are slow due to poor caching values. Could be due to poorly configured authentication.

qxd 2/25/05 10:19 AM Page 475 Other NFS Considerations 475 Security NFS and portmap have had a number of known security deficiencies in the past. the NFS client waits indefinitely for it to return. You can exercise security-related caution with NFS by following a few guidelines: ☞ Restrict its use to secure networks. use the soft option in the NFS client’s /etc/fstab file. ☞ Do not have mission-critical computers rely on an NFS server to operate. ☞ Consider using read-only exports whenever data updates aren’t necessary. however. the list. ☞ Always use the sync option whenever possible. unless the server’s reliability can be guaranteed. especially if the NFS client and NFS server are being managed by different sets of administrators. These points should be the foundation of your NFS security policy. This is normally an undesirable condition. NFS Hanging As stated before. . ☞ Do not include NFS-mounted directories as part of your search path. This also forces programs relying on the same client server relationship to wait indefinitely too. because a hung NFS connection to a directory in your search path could cause your shell to pause at that point in the search path until the NFS session is regained. NFS doesn’t encrypt data and it is possible for root users on NFS clients to have root access to the server’s filesystems. This causes NFS to report an I/O error to the calling program after a long timeout. ☞ Use the root_squash option in /etc/exports file (default) to reduce the risk of the possibility of a root user on the NFS client having root file permission access on the NFS server. I don’t recommended using NFS over insecure networks. I’d suggest that you refer to a dedicated NFS reference for more detailed advice.32Harrison_ch29. isn’t comprehensive due to the concise scope of this book. ☞ Export only the most needed data. For this reason. You can reduce the risk of NFS hanging by taking a number of precautions: ☞ Run NFS on a reliable network. As a result. ☞ Avoid having NFS servers that NFS mount each other’s filesystems or directories. if the NFS server fails.

This can be disabled with the no_root_squash export option in the /etc/exports file. therefore. Test your network-based applications thoroughly before considering using NFS. Limiting root Access NFS doesn’t allow a root user on a NFS client to have root privileges on the NFS server.qxd 2/25/05 10:19 AM Page 476 476 Remote Disk Access With NFS Chapter 29 Following these simple but important guidelines will help to make NFS operate more predictably. intermittent or insecure. but the data is frequently transferred on a randomly chosen TCP port negotiated between them. This is normally an undesirable condition. . It is best. File Locking NFS allows multiple clients to mount the same directory. to use the /etc/exports file to limit access to certain trusted servers or networks. If a user named nfsuser exists on the NFS client. although more recent versions are said to have rectified the problem. then they will have access to all the files of a user named nfsuser on the NFS server. Restricting Access to the NFS Server NFS doesn’t provide restrictions on a per-user basis. but that is beyond the scope of this book. File Permissions The NFS file permissions on the NFS server are inherited by the client. allowing you more time to focus on other important administrative tasks. but NFS has a history of not handling file locking well.32Harrison_ch29. You may also want to use a firewall to protect access to the NFS server. Nesting Exports NFS doesn’t allow you to export directories that are subdirectories of directories that have already been exported unless they are on different partitions. A main communication control channel is usually created between the client and server on TCP port 111. You may also want to eliminate any wireless networks between your NFS server and client. especially if the NFS client and NFS server are being managed by different sets of administrators. and it is not wise to mount an NFS share across the Internet as access could be either slow. It can become complicated especially if the users and user groups on the NFS client that are expected to access data on the NFS server don’t exist on the NFS server. There are ways to limit the TCP ports used.

such as a database stored on a centralized server. . Many of the new network-attached storage products currently available on the market rely on NFS—a testament to its popularity. CONCLUSION As you have seen NFS can be a very powerful tool in providing clients with access to large amounts of data. make the key users and groups on both systems match and make sure the permissions on the NFS client mount point and the exported directories of the NFS server are in keeping with your operational objectives.32Harrison_ch29. and improving security.qxd 2/25/05 10:19 AM Page 477 Conclusion 477 For simplicity. increasing stability.

qxd 2/25/05 10:19 AM Page 478 .32Harrison_ch29.

qxd 2/25/05 10:20 AM Page 479 C H A P T E R 30 Centralized Logins Using NIS IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Scenario Configuring the NFS Server Configuring the NFS Client Configuring the NIS Server Adding New NIS Users Configuring the NIS Client NIS Slave Servers Changing Your NIS Passwords Considerations for a Non-NFS Environment NIS Troubleshooting Conclusion Network Information Services (NIS) enables you to create user accounts that can be shared across all systems on your network. An advantage of NIS is that users need to change their passwords on the NIS server only. This makes NIS popular in computer training labs. 479 . NIS clients download the necessary username and password data from the NIS server to verify each user login. but I suggest that you restrict its use to highly secure networks or to networks where access to non-NIS networks is highly restricted. distributed software development projects or any other situation where groups of people have to share many different computers.33Harrison_ch30. instead of every system on the network. A detailed analysis of NIS security is beyond the scope of this book. The disadvantages are that NIS doesn’t encrypt the username and password information sent to the clients with each login and that all users have access to the encrypted passwords stored on the NIS server. The user account is created only on the NIS server.

4. ☞ Each user’s home directory will be automatically mounted with each user login on the PCs using NFS. “Installing RPM Software.sync) /home directory . The lab instructor did some research and created an implementation plan: 1. CONFIGURING THE NFS SERVER Here are the steps to configure the school’s NFS server: 1. A school wants to set up a small computer lab for its students: ☞ The main Linux server. Configure Smallfry as an NIS client. Convert the account to a NIS user account. 6. 2. Downloading and installing RPMs isn’t hard. Configure Bigboy as an NFS server to make its /home directory available to the Linux workstations.” When searching for the RPMs.i386. as discussed in Chapter 6. it’s time to get to work. remember that the filename usually starts with the software package name followed by a version number. consider an example. 5. Test a remote login from Bigboy to Smallfry using the username and password of the account nisuser. Configure Smallfry as an NFS client that can access Bigboy’s /home directory.33Harrison_ch30. ☞ Users logging into the PCs will be assigned home directories on Bigboy and not on the PCs themselves. Configure Bigboy as an NIS server. ☞ The lab instructor will practice with a Linux PC named Smallfry before implementing NIS on all the remaining PCs. 3.rpm. You have the scenario and the plan. Create a user account (nisuser) on Bigboy that doesn’t exist on Smallfry. as in yp-tools-2. and ypbind and yp-tools are on the client.qxd 2/25/05 10:20 AM Page 480 480 Centralized Logins Using NIS Chapter 30 SCENARIO To understand the benefits of NFS. ☞ The suite of NIS RPMs have been installed on the server and client: ypserve and yp-tools are on the server. Edit the /etc/exports file to allow NFS mounts of the with read/write access: /home *(rw. has a large amount of disk space and will be used as both the NIS server and NFS-based file server for the Linux PCs in the lab.8-3. Bigboy.

and portmap daemons are running and configured to start after the next reboot: [root@smallfry tmp]# chkconfig nfslock on [root@smallfry tmp]# chkconfig netfs on [root@smallfry tmp]# chkconfig portmap on [root@smallfry tmp]# service portmap start Starting portmapper: [ OK ] [root@smallfry tmp]# service netfs start Mounting other filesystems: [ OK ] [root@smallfry tmp]# service nfslock start Starting NFS statd: [ OK ] [root@smallfry tmp]# .qxd 2/25/05 10:20 AM Page 481 Configuring the NFS Client 481 2. you’d have to force the users to log off. 1. these prerequisites aren’t necessary. This will be covered next. These steps archive the /home directory. nfslock. Make sure the required netfs. As this is a lab environment. we have to configure its clients. back up the data. restore it to the NFS server. In a production environment in which the /home directory would be actively used. CONFIGURING THE NFS CLIENT You also need to configure the NFS clients to mount their /home directories on the NFS server. and portmap daemons are both running and configured to start after the next reboot. Let NFS read the /etc/exports file for the new entry. nfslock.33Harrison_ch30. [root@bigboy tmp]# chkconfig nfslock on [root@bigboy tmp]# chkconfig nfs on [root@bigboy tmp]# chkconfig portmap on [root@bigboy tmp]# service portmap start Starting portmapper: [ OK ] [root@bigboy tmp]# service nfslock start Starting NFS statd: [ OK ] [root@bigboy tmp]# service nfs start Starting NFS services: [ OK ] Starting NFS quotas: [ OK ] Starting NFS daemon: [ OK ] Starting NFS mountd: [ OK ] [root@bigboy tmp]# After configuring the NIS server. Make sure the required nfs. and then follow the steps below. and make available to the network with the exportfs command: [root@bigboy tmp]# exportfs -a [root@bigboy tmp]# /home 3.

After five minutes.home --timeout 600 5.1.qxd 2/25/05 10:20 AM Page 482 482 Centralized Logins Using NIS Chapter 30 2.. Start autofs and make chkconfig command: sure it starts after the next reboot with the [root@smallfry tmp]# chkconfig autofs on [root@smallfry tmp]# service autofs restart Stopping automount:[ OK ] Starting automount:[ OK ] [root@smallfry tmp]# After doing this... .33Harrison_ch30.home for mounting information whenever the /home directory is accessed. Start configuring autofs automounting.1..168. [root@smallfry tmp]# mv /home /home.home to do the NFS mount whenever the /home directory is accessed.100:/home /home/ ls /home quotauser smallfry www umount /home 4. If the line is too long to view on your screen.save .home * -fstype=nfs. Unmount it once everything looks correct: [root@smallfry tmp]# [root@smallfry tmp]# ftpinstall nisuser [root@smallfry tmp]# [root@smallfry tmp]# mount 192..tcp \ 192..intr. Make sure you can mount Bigboy’s /home directory on the new /home directory you just created.master /home /etc/auto...100:/home:& 6.168. and create a new directory /home on which you’ll mount the NFS server’s directory.wsize=8192.rsize=8192. drwxr-xr-x 1 root root 11 Nov 16 20:22 home drwxr-xr-x 2 root root 4096 Jan 24 2003 home.master file to refer to file /etc/auto. . autofs unmounts the directory: #/etc/auto.soft.nosuid. Keep a copy of the old /home directory.save [root@smallfry tmp]# mkdir /home [root@smallfry tmp]# ll / . Edit file /etc/auto. you won’t be able to see the contents of the /home directory on Bigboy as user root. you can add a \ character at the end to continue on the next line: #/etc/auto. You’ll be able to test this later after NIS is configured. Edit your /etc/auto. . This is because by default NFS activates the root squash feature. which disables this user from having privileged access to directories on remote NFS servers. [root@smallfry tmp]# 3.

” for details. NIS was called Yellow Pages. All newly added Linux users will now be assigned a home directory under the new remote /home directory.0. “Remote Disk Access With NFS. when in reality they are automatically mounted and accessed over your network. yet many of the key programs associated with NIS have kept their original names beginning with yp. The configuration of the NIS server is not difficult.0.conf File NIS servers also have to be NIS clients themselves. See Chapter 29. For the school. but requires many steps that you may overlook. so you’ll have to edit the NIS client configuration file /etc/yp. call the domain NIS-SCHOOL-NETWORK: #/etc/sysconfig/network NISDOMAIN=”NIS-SCHOOL-NETWORK” Edit Your /etc/yp.qxd 2/25/05 10:20 AM Page 483 Configuring the NIS Server 483 Note This automounter feature doesn’t appear to function correctly in my preliminary testing of Fedora Core 3. The developers had to change the name after a copyright infringement lawsuit. You now have to configure NIS login authentication for the lab students before the job is done.conf . Don’t worry. CONFIGURING THE NIS SERVER NFS only covers file sharing over the network. Note In the early days. we’ll review each one in detail.conf to list the domain’s NIS server as being the server itself or localhost: # /etc/yp.1 .33Harrison_ch30. This scheme will make the users feel their home directories are local. Edit Your /etc/sysconfig/network File You need to add the NIS domain you wish to use in the /etc/sysconfig/network file.ypbind configuration file ypserver 127.

You’ll start these daemons after initialization is completed.33Harrison_ch30.1 Required NIS Server Daemons Purpose The foundation RPC daemon upon which NIS runs Lets users change their passwords on the NIS server from NIS clients Main NIS server daemon Main NIS client daemon Used to speed up the transfer of very large NIS maps Daemon Name portmap yppasswdd ypserv ypbind ypxfrd Make sure they are all running before continuing to the next step.d directory and use the chkconfig command to ensure they start after the next reboot. You can use the rpcinfo command to do this: [root@bigboy tmp]# rpcinfo -p program vers proto port 100000 2 tcp 111 100000 2 udp 111 100009 1 udp 681 100004 2 udp 698 100004 1 udp 698 100004 2 tcp 701 100004 1 tcp 701 [root@bigboy tmp]# localhost portmapper portmapper yppasswdd ypserv ypserv ypserv ypserv The ypbind and ypxfrd daemons won’t start properly until after you initialize the NIS domain. Table 30.qxd 2/25/05 10:20 AM Page 484 484 Centralized Logins Using NIS Chapter 30 Start the Key NIS Server-Related Daemons Start the necessary NIS daemons in the /etc/init.1 lists a summary of the daemon’s functions. . [root@bigboy tmp]# service portmap start Starting portmapper: [ OK ] [root@bigboy tmp]# service yppasswdd start Starting YP passwd service: [ OK ] [root@bigboy tmp]# service ypserv start Setting NIS domain name NIS-SCHOOL-NETWORK: Starting YP server services: [ OK ] [root@bigboy tmp]# [root@bigboy [root@bigboy [root@bigboy [root@bigboy tmp]# chkconfig portmap on tmp]# chkconfig yppasswdd on tmp]# chkconfig ypserv on tmp]# [ OK ] Table 30.

.bynumber..byaddr. Updating protocols...byname.... gmake[1]: Entering directory `/var/yp/NIS-SCHOOL-NETWORK’ Updating passwd. Now you can run ypinit -s bigboy on all slave server.. Running /var/yp/Makefile. With this procedure. such as: failed to send ‘clear’ to local ypserv: RPC: Port mapper failureUpdating group... Updating group. You will be prompted for the name of the NIS server.bygid.byname. bigboy is in the list of NIS server hosts. type a <control D>....byname. Building /var/yp/NIS-SCHOOL-NETWORK/ypservers. Updating group. Updating passwd...aliases.. Updating hosts. we have to construct a list of the hosts which will run NIS servers.byuid. Updating rpc.. Updating hosts. ....byname.. Updating mail.byname.byname. which in this case is Bigboy.33Harrison_ch30.. Updating services.. Updating protocols. Updating netid... Please continue to add the names for the other hosts. Updating services.qxd 2/25/05 10:20 AM Page 485 Configuring the NIS Server 485 Initialize Your NIS Domain Now that you have decided on the name of the NIS domain.bygid. [root@bigboy tmp]# Note Make sure portmap is running before trying this step or you’ll get errors.... next host to add: bigboy next host to add: The current list of NIS servers looks like this: bigboy Is this correct? [y/n: y] y We need a few minutes to build the databases. gmake[1]: Leaving directory `/var/yp/NIS-SCHOOL-NETWORK’ bigboy has been set up as a NIS master server.byname....bynumber..byservicename. Updating rpc. When you are done with the list. all nonprivileged accounts are automatically accessible via NIS: [root@bigboy tmp]# /usr/lib/yp/ypinit -m At this point. one per line.. you’ll have to use the ypinit command to create the associated authentication files for the domain....

yppasswd. [root@bigboy tmp]# service ypxfrd start Starting YP map server: [ OK ] [root@bigboy tmp]# chkconfig ypbind on [root@bigboy tmp]# chkconfig ypxfrd on Make Sure the Daemons Are Running All the NIS daemons use RPC port mapping and.qxd 2/25/05 10:20 AM Page 486 486 Centralized Logins Using NIS Chapter 30 You will have to delete the /var/yp/NIS-SCHOOL-NETWORK directory and restart portmap. Start the ypbind and ypxfrd Daemons You can now start the ypbind and the ypxfrd daemons because the NIS domain files have been created: [root@bigboy tmp]# service ypbind start Binding to the NIS domain: [ OK ] Listening for an NIS domain server.33Harrison_ch30. are listed using the rpcinfo command when they are running correctly. [root@bigboy tmp]# rpcinfo -p localhost program vers proto port 100000 2 tcp 111 portmapper 100000 2 udp 111 portmapper 100003 2 udp 2049 nfs 100003 3 udp 2049 nfs 100021 1 udp 1024 nlockmgr 100021 3 udp 1024 nlockmgr 100021 4 udp 1024 nlockmgr 100004 2 udp 784 ypserv 100004 1 udp 784 ypserv 100004 2 tcp 787 ypserv 100004 1 tcp 787 ypserv 100009 1 udp 798 yppasswdd 600100069 1 udp 850 fypxfrd 600100069 1 tcp 852 fypxfrd 100007 2 udp 924 ypbind 100007 1 udp 924 ypbind 100007 2 tcp 927 ypbind 100007 1 tcp 927 ypbind [root@bigboy tmp]# . therefore. and ypserv before you’ll be able to do this again successfully.

nonprivileged accounts become automatically accessible via NIS.. In this case.byuid..byname. ypmatch. it just provides the user’s entry in the /etc/passwd file.. which has similar syntax. getent [root@bigboy yp]# getent passwd nisuser nisuser:x:504:100::/home/nisuser:/bin/bash [root@bigboy yp]# . Updating passwd.. Once this is complete. [root@bigboy tmp]# cd /var/yp [root@bigboy yp]# make gmake[1]: Entering directory `/var/yp/NIS-SCHOOL-NETWORK’ Updating passwd.byname. such as the login shell. New password: Retype new password: passwd: all authentication tokens updated successfully. and home directory: [root@bigboy tmp]# useradd -g users nisuser [root@bigboy tmp]# passwd nisuser Changing password for user nisuser. On a NIS client. the results are identical with both showing the encrypted password. Updating netid. you’ll create a user account called nisuser and give it a new password. which should return the user’s encrypted password string: [root@bigboy yp]# ypmatch nisuser passwd nisuser:$1$d6E2i79Q$wp3Eo0Qw9nFD/::504:100::/home/nisuser:/bin/bash [root@bigboy yp] You can also use the getent command. This procedure makes all NIS-enabled. you then have to update the NIS domain’s authentication files by executing the make command in the /var/yp directory. It also exports all the user’s characteristics stored in the /etc/passwd and /etc/group files. not just newly created ones. gmake[1]: Leaving directory `/var/yp/NIS-SCHOOL-NETWORK’ [root@bigboy yp]# You can check to see if the user’s authentication information has been updated by using the ypmatch command. the user’s group.qxd 2/25/05 10:20 AM Page 487 Adding New NIS Users 487 ADDING NEW NIS USERS New NIS users can be created by logging into the NIS server and creating the new user account.33Harrison_ch30... Unlike doesn’t provide an encrypted password when run on an NIS server.

It also edits the /etc/sysconfig/network file to define the NIS domain to which the NIS client belongs. . amongst other things. the IP address of the NIS server for a particular domain.168. nsswitch. and NIS.100 #/etc/sysconfig/network NISDOMAIN=NIS-SCHOOL-NETWORK In addition.ypbind configuration file domain NIS-SCHOOL-NETWORK server 192. Run authconfig The authconfig program automatically configures your NIS files after prompting you for the IP address and domain of the NIS server: [root@smallfry tmp]# authconfig Once finished.33Harrison_ch30.1.qxd 2/25/05 10:20 AM Page 488 488 Centralized Logins Using NIS Chapter 30 CONFIGURING THE NIS CLIENT Now that the NIS server is configured. There are a number of related configuration files that you need to edit to get it to work.conf passwd: files nis shadow: files nis group: files nis Note You can also locate a sample NIS yp-tools* directory. Remember to use the rpcinfo command to ensure they are running correctly. # /etc/yp. LDAP.d directory and use the chkconfig command to ensure they start after the next reboot. the authconfig program updates the /etc/nsswitch. yppasswd.conf file in the /usr/share/doc/ Start the NIS Client-Related Daemons Start the ypbind NIS client.conf file that lists the order in which certain data sources should be searched for name lookups. and portmap daemons in the /etc/init. such as those in DNS.conf .conf file that defines. Take a look at the procedure. Here you can see where NIS entries were added for the important login files: #/etc/nsswitch. it should create an /etc/yp. it’s time to configure the NIS clients.

or more simply by editing the /etc/hosts file on both Linux boxes: # # File: /etc/hosts (smallfry) # 192..L0XqYJUk09Bw.pJ/:504:100::/home/www:/bin/bash smallfry:$1$qHni9dnR$iKDs7gfyt. [root@smallfry tmp]# service yppasswdd start Starting YP passwd service: [ OK ] [root@smallfry tmp]# [root@smallfry tmp]# chkconfig ypbind on [root@smallfry tmp]# chkconfig portmap on [root@smallfry tmp]# chkconfig yppasswdd on Remember to use the all started correctly.168.:501:100::/:/bin/bash [root@smallfry tmp]# [root@smallfry tmp]# ypmatch nisuser passwd nisuser:$1$d6E2i79Q$wp3Eo0Qw9nFD/:504:100::/home/nisuser:/bin/bash [root@smallfry tmp]# . Verify Name Resolution rpcinfo -p localhost command to make sure they As the configuration examples refer to the NIS client and server by their hostnames.1.:502:100::/:/bin/bash www:$1$DDCi/OPI$hwiTQ.33Harrison_ch30. This can be configured either in DNS.168.qxd 2/25/05 10:20 AM Page 489 Configuring the NIS Client 489 [root@smallfry tmp]# service portmap start Starting portmapper: [ OK ] [root@smallfry tmp]# service ypbind start Binding to the NIS domain: Listening for an NIS domain server.1. [root@smallfry tmp]# ypcat passwd nisuser:$1$Cs2